^^
Origin of Christian
CHURCH ART
. Oxford University Press
London Edinburgh Glasgow Copenhagen
New York Toronto Melbourne Cape Town
Bombay Calcutta Madras Shanghai
Humphrey Milford Publisher to the University
Earls Barton Tower, Northamptonshire.
See p. 234.
Origin of Christian
CHURCH ART
New Facts and Principles
of Research
BY
JOSEF STRZYGOWSKl
Sight Lectures delivered for the Olaus-Petri
Foundation at Upsaia, to which is added
a chapter on Christian oirt in "Britain
TRANSLATED FROM THE GERMAN
BY
O. M. DALTON, M.A, and H.J. BRAUNHOLTZ, M.A.
OXFORD
AT THE CLARENDON PRESS
1923
4 800
705fi:i3
Printed in England
TO
MY FRIENDS
IN
OXFORD
PREFACE
THERE may be not a few who would care to know how the
problem of the origin of Christian art (I purposely exclude
sepulchral art from the discussion) presents itself to-day
to an investigator with more than thirty years of unremitting
labour in the East behind him. That no claim to finality can
be made will be readily understood by all who reject easy move-
ment along the ruts prepared by some chosen School, preferring
to break their own way through obstacles to the truth. The aim
of the present book is merely to summarize conclusions reached
after the personal research and individual struggle of decades.
Many gulfs are yet unbridged ; the gaps often seem broader
than the firm ground, but it is worth while to survey the
achievement as a whole, to bring isolated facts into their true
relation to guiding ideas, and to point out the paths which the
next generations will probably have to tread. Each of my eight
chapters is concerned with such a path.
The first chapter estabUshes a new horizon extending from
Europe, the Mediterranean and the ancient East, as far as Persia.
The Iranian people, the second great source of Aryan energy,
joined forces with the inhabitants of the North and the pastoral
nomadic tribes to develop the ' mediaeval ' spirit in Christian art,
and in the South had already made it prevail before Northern
Europe laid hand to the work.
The second chapter follows the successive changes in
social order in the early stages of Christian art. It traces the
development from Founder and primitive community to State
and Church ; it shows the artist drawn into the service of
autocracy ; and finally asks how much was now left for the
initiative bom of artistic freedom.
The third chapter seeks to show how in the fourth century
East- Aryan Christianity advanced by way of Armenia, bringing
3451
x>;
viii PREFACE
with it the centralized domical plan. It further attempts to
explain how during the same period the Semites of Mesopotamia
introduced their long barrel-vaulted churches among the timber-
roofed basilicas of the Mediterranean area.
The fourth chapter describes the * styles ' which succeeded
that of Early Christian art in the West : Romanesque, Gothic,
Renaissance. It reviews in chronological sequence elements which
we have already seen existing side by side in the fourth century.
This is a theme which I shall expand in Beitrdge zur vergleichen-
den Kunstforschung (Introduction to Part I).
The fifth chapter conveys a warning against the prevalent
over-estimate of representational art. The non-representational
art of Islam must have had a prototype in that of Mazdaism, the
whole principle of which, as a creation of the North, was antago-
nistic to the Southern spirit, as I explain in my Altai-Iran und
Volkerwanderung .
The sixth endeavours to prove that at the beginning, and
even as late as the fourth century. Christian art was non-
representational in the decoration of its churches. It shows how
precipitate it is, on the mere evidence of the Sarcophagi and
the mural paintings in the Catacombs, to pronounce Early
Christian a branch of the classical stem. Only the supposition
of an original non- representational system can render the later
iconoclastic disturbances intelligible.
In the seventh chapter I try to explain the reasons for the
outbreak against pictures. Semitic art, outbidding that of
Greece, had begun to personify all things and sundry ; it trans-
ferred to the sphere of Christianity the spell-binding methods
of representation which a pagan despotism had first devised.
It was opposed by the ancient Aryan concept of the universe,
which we may still see shedding the light of Northern magic in
the South through landscape-mosaics which do not attempt to
render nature.
In the eighth chapter I separate out the purely artistic
results reached in the preceding sections, classify them, and
point out the gaps which I detect. I endeavour to convince the
PREFACE ix
reader that archaeology must give up its false methods, the
philological and historical, based on texts or the chance survival
of individual monuments, and the philosophical and aesthetic,
which evades the facts of evolution. The history of art must
work itself free from the mere comparative study of monuments ;
it must concentrate upon the work of art and its values, absolute
and evolutional, and so find a path of its own. Only then can it
take its proper place among the sister sciences older than itself.
I hope to pursue this subject further in my System und Methode
der Kunstforschung.
I have throughout avoided questions of the day, whether
political or religious. Names are rarely mentioned. I occupy
myself rather with opinions expressed on new facts and principles
which, as experience shows, will yet be contested for decades
to come. This book marks a pause, it is retrospective ; the
writer trusts that it may elicit the criticism of individuals or
groups of scholars in general agreement with his line of thought,
though not pursuing it from the artistic standpoint alone. But
there will remain a few representatives of art-history with whom
he can hardly hope ever to reach agreement.
Illustration has only been employed where it seemed really
to further the general understanding of the subject. The book
has been kept within limits by substituting for references in
notes a bibliography at the end, including both my own works
and those of my colleagues ; here the reader will find references
in abundance alike to monuments and to texts. Both in biblio-
graphy and illustration a knowledge of the Western material
and its literature is assumed ; but it is much to be desired that
any prospective collaborators in research should begin by com-
pletely mastering the works mentioned on pp. 4, 5 and 6, above
all Altai-Iran und Volkerwanderung , and Die Baukunst der
Armenier und Europa. What happens when this is not done is
only too clearly seen in certain * reviews ' of the last-mentioned
book.
The eight lectures forming the basis of these chapters were
written in the country during the summer of 19 18, and delivered
X PREFACE
in the University of Upsala, March 24 to April 7, 1919, with the
accompaniment of lantern sUdes. The method of oral delivery,
with its opportunities of enforcing the emphasis at the right points,
was naturally of service to the writer ; it helped him to sift and
order his leading ideas. He was consequently desirous of
revising his work in accordance with the experience thus gained.
But there were obstacles to such a course. Firstly, the Swedish
translation was completed ; secondly, throughout the winter of
1918-19 the writer had been occupied almost exclusively with
the significance of the North for the artistic development of
Europe ; the new subject thrust the old aside, and the material
of these lectures was already less vividly present to his mind.
All that was possible was to insert in the two last chapters of
this, the original German edition, a summary and survey written
after the lectures at Upsala had been delivered.
Josef Strzygowski.
Vienna.
October 1919.
TRANSLATORS' PREFACE
The translators have not found it easy to produce a satis-
factory rendering of Ursprung der Christlichen Kirchenkunst.
The style of the original does not lend itself readily to an English
version, and it has been almost impossible to find concise equiva-
lents for some of the terms employed, A few short passages of
purely ephemeral reference have been omitted ; otherwise the
book remains substantially as it was written. The proofs have
passed through the author's hands.
London, 1923.
TABLE OF CONTENTS
• I. THE NEW HORIZON
1. Geographical Influences
2. Historical Influences .
3. Social Conditions
4. Christianity in Persia .
5. The Asiatic East
PAGE
I
7
10
16
23
^U. COMMUNITY, CHURCH, AND COURT
1. Christian Popular Art to A. D. 400
2. Church Art of the Fifth Century .
3. Autocratic Procedure of the Court
4. Santa Sophia . . ,
28
30
35
42
46
rf^III. LOCAL VARIATION IN THE EARLY CHRISTIAN
VAULTED ARCHITECTURE OF THE EAST
1. The Dome .......
A. The many-domed type . . .
B. The one-domed type .....
a. Subdivisions of the niche-buttressed square
b. Buildings composed of niche-buttresses only
2. The Barrel Vault
A. The church with transverse nave
B. The long church
3. Abutment of the Domed Building
A. The trefoil (Trikonchos)
B. The domed basilica
C. The cruciform domed church
D. The domed hall-church
51
57
60
6i
61
64
67
67
68
69
69
70
71
7a
xu
TABLE OF CONTENTS
IV. THE SUCCESSION OF PERIODS IN WESTERN
ARCHITECTURE
1 . The Wooden Churches of the North .
2. The Oriental Art of Western Europe (Romanesque)
A. Barrel-vaulted Churches ....
a. Barrel-vaulted Churches with single nave
b. Hall-Churches .....
B. Churches with several Domes
C. The Groined Vauh
3. The Northern Art of Europe (Gothic) .
4. The ItaHan Mixed Style (Renaissance) .
PAGE
75
78
81
86
87
88
89
90
93
97
V. RELIGIONS WITHOUT REPRESENTATIONAL ART 102
I. Islam .......
4
105
2. Iran .......
108
3. Mazdaism
115
A. The Hvarenah Landscape .
118
B. Hvarenah Symbols
120
C. Decoration of the Mazdean House
. 124
D. Mazdean Costume
. 125
4. Manichaean Art ....
. 126
VI. NON-REPRESENTATIONAL CHURCH ART, AND
THE SUBSEQUENT ANTI -REPRESENTA-
TIONAL MOVEMENT 130
1 . Italian Mosaics ......
A. The Decoration of the Dome
B. Barrel- vault Decoration
C. Apse Decoration ....
2. The Eastern ' Hinterlands ' of the Mediterranean
3 . The East- Aryan Province ....
A. Structural Members ....
a. The Support ....
b. Door and Gate ....
133
133
134
136
138
144
145
145
147
TABLE OF CONTENTS
xui
B. Decoration of flat surfaces
a. Crosses
b. Landscapes .
c. Scenes from the Chase
)^ d. Vine Scrolls
e. Dawn clouds
C. Woven and embroidered Stuffs
4. The European North .
PAGE
148
149
149
150
150
151
152
VI L THE TRIUMPH OF REPRESENTATIONAL ART
HELLENISM, SEMITISM, MAZDAISM
Symbolic representation of the West- Aryans
Semitic Realism ....
A. The Didactic Element
B. Aramaean Cycle
C. Sequence of Didactic Subjects
Mazdean Ideas in Christian Representation
A. The Mounted Saint .
B. The Last Things
C. The Last Judgement .
D. The mosaic of SS. Cosmas and Damian
Influence of the Court
The West
Handicraft ....
1 . Material and Work
A. Building Material
B. Work .
Spiritual Values
Significance .
2. Subject
y A. Non-representational Decoration
B, Representational Decoration .
155
158
161
163
166
169
171
173
176
176
178
182
184
VIII. SYSTEMATIC INVESTIGATION OF ESSENTIAL
CHARACTER AND APPLICATION OF THE
COMPARATIVE METHOD 189
193
194
194
196
198
199
200
202
204
xiv TABLE OF CONTENTS
PAGB
Appearance ......... 206
3. Shape . 207
A. Constructional Form ..... 208
B. Decoration ....... 209
C. Representation . . . . . .210
4. Design 212
A. Mass . . . . . . . . 213
B. Space . . . . . . . 216
C. Light ........ 219
D. Colour ....... 220
5. Content 221
A. Intercession ...... 222
B. Instruction ....... 223
C. Use of Dazzling Effects .... 223
/ D. Artists ....... 224
IX. HIBERNO-SAXON ART IN THE TIME OF BEDE . 230
1 . The Study of the Monuments ..... 230
2. Investigation of Essential Character .... 233
(i) Material and Work ...... 233
(2) Object or Purpose . . . . . . 236
(3) Shape .238
(4) Form 242
(5) Content . . . . . . . . 246
3. History of Development ...... 246
4. The Subjective Side ....... 250
BIBLIOGRAPHY . .253
INDEX 261
LIST OF ILLUSTRATIONS
FIG. FACmc PAGE
Earls Barton Tower, Northamptonshire ; photo. J. R H. Weaver Frontispiece
J. Ma^tara, domed square with apse-buttresses {c. A. d. 650) ; exterior from
south-west ........... 61
62
62
2. Artik, cathedral, domed square with apse-buttresses ; south-west view
3. Artik, cathedral ;' plan ........
4. Ani, Church of S. Gregory (mid-tenth century), sexfoil plan with triangu
lar slits ; south-east view .......
5. Zwa rthnotz, Church of S. Gregory (c. A. D. 650), quatrefoil plan with
ambulatory, and palace ; plan ......
6. Salah, Mar Yakub, church with transverse nave ; photo. G. L. Bell
7. Salah, Mar Yakub ; plan ; after G. L. Bell
8. Ereruk, basilica with ruined barrel- vaults, and fa9ade towers ; interior
See also fig. 21 ........ .
9. Ereruk, basilica ; plan. See also fig. 21 .
10. Thalin, cathedral (sixth-seventh century), from south-west ; three-aisled
trefoil-ended type ; the dome fallen in
11. Ani, cathedral (before A. D. 989-1001); long cruciform domed type
north-east view .........
66
66
68
68
69
69
70
71
12. Thalish (a. D. 668), domed hall-church, south-east side ; the dome fallen in 72
13. Thalish, the same ; interior from west ...... 73
14. Valdres, typical wooden church ; plan ; after Dietrichsen ... 80
15. Borgund, wooden church ; plan; after Dietrichsen .... 80
16. Borgund, wooden church ; interior ; after Dietrichsen .... 80
17. Triforium of wooden churches ; after Dietrichsen . . . .80
18. Salah, Church of S. James, type with transverse nave ; vaulting of nave
photo. G. L. Bell .......... 91
19. Ereruk; south view. See also figs. 10 and 11 ..... 91
20. Sofia, Church of S. Sophia, cruciform domed church with three aisles
plan ; after Filow 91
21. Ani, cathedral ; interior from north-west 94
22. Ani, cathedral ; plan . 94
23. Ani, cathedral ; bases of piers with clustered columns .... 95
24. Thalin, cathedral ; deuil of south-west pier 95
25. Aisasi, porch of monastery church ; Armenian ribbed vaulting . . 96
26. Steyr, court of a house ; photo. ReiflFenstein 97
xvi LIST OF ILLUSTRATIONS
FIG. FACING PAGE
VJ. Vagharshapat, Hripsimeh (a. d. 6i8), domed square with apse-buttresses
and angle-chambers ; south view ....... 98
28. Vagharshapat, Hripsimeh ; plan ....... 98
29. Amman, square building with apse-buttresses ; blind arcade with formal
ornament . . . . . . . . . .99
30. Ani, church of the Apostles ; plan ....... 99
31. Korea, stone tomb ; structure of the roof ; from the A^oAAa ... 99
32. Samath, stupa ; zone of ornament; photo. E. La Roche . . .114
33. Island of Achthamar, Lake Van, cruciform church ; south view . • "4
34. Mshatta, facade ; to left, end of the Hvarenah wall-facing . . • "5
35. Kairwan, wooden mimbar ; pierced interlacing, niches, &c. . . • "7
36. Brunswick Museum, ivory carving ; Yima ...... 120
37. Bazaklik, Chinese Turkestan, temple group ; painting ; Padmapani on
scrolled stem ; after A. Griinwedel 120
38. Spoleto, fa9ade of S. Pietro ; photo. Alinari . . . . . .121
39. Island of Achthamar, Lake Van, cruciform church (a. d. 915-921) ;
east side ........... 122
40. Lemberg, Armenian gospel of A. D. 1 198 ; Eusebian canons . . , 123
41 . Vienna, Austrian Folk-Museum ; Armenian silk embroidery from
Bukovina ........... 124
42. Sorcuq, Chinese Turkestan ; fresco in the Naga Cave ; after A. Griin-
wedel . . . . . . . . . . . 137
43. Oasis of El-Khargeh ; apse decoration of a funerary chapel ; after
V. de Bock ........... 137
44. Angora, Asia Minor ; so-called column of Augustus .... 146
45-7. Bisutun and Ispahan ; capitals . . , . . . .146
48. Edessa ; impost capital ......... 146
49. Baghdad ; impost capital . . . . . . . 146
50. Meiafarqin, church of the Virgin; interior from west door; photo.
G. L. Bell 147
51. Mshatta, capital from the trefoil-ended hall (Trikonchos) ; left half . 147
52. Dara ; basket capital ......... 147
53. Armenia, Ketcharus monastery ; south door of Church of S. Gregory . 152
54. Dashlut, Egypt ; door of Court of Mosque ; the mounted saint, half
destroyed ........... 152
55. Vienna, in private possession ; woollen fabric from Egypt . , . 153
56. Urnas, wooden church ; door with interlaced beasts ; after A. Haupt . 153
57. Island of Achthamar, Lake Van, cruciform church ; west side . '159
58. Voronetz, Bukovina, monastery church ; south-west view . . .160
59. Dara, rock cave ; exterior ......... 166
60. Berlin, Kaiser Friedrich Museum ; mounted Christ from Upper Egypt . 166
61. 62. Florence, Laurentian Library ; MS. of Rabula (a. d. 586) ; Eusebian
Canons ; photo. G. Millet ........ 167
LIST OF ILLUSTRATIONS
xvii
FIG. FACING
63. Rome, SS. Cosmas and Damian ; apse-mosaic; after Wilpert
64. Kioto, Zenrinji temple ; figured silk ; after the Kokka .
65. Odzun ; pillars from monument .....
66. Acca Cross .........
67. Ruthwell and Bewcastle Crosses ; reproduced by permission of John
Murray from Prof. Baldwin Brown's The Arts in Early England
68. Acre pillars in Venice ......
69. ' Nigg-stone '........
70. Stucco ornaments from the Deir es-Suryani
71. Page of the Lindisfarne Gospels, f. 2c'' ; British Museum
72. Mzchet ; the Christ outside the apse ....
73. Bradford-on-Avon ; interior; reproduced by permission of the Syndics
of the Cambridge University Press from Sir T. G. Jackson's Byzantine
and Romanesque Architecture ........
74. Heading of MS. Paris Bibl. Nat. 543, f. 28"
PAGE
1 80
l8l
238
240
241
242
244
24s
246
247
I
The New Horizon
ART translates inward meaning into visible form ; it uses
/-\ the creative skill of man to free it from the limitations of
■*^ ^ life. Under its religious aspect it involves a compromise
between the artist and the visible world, due to the yearning of
the human soul for positive belief. Christian art receives its
name from a Founder whose life was so intent on an invisible
world that neither construction nor presentation of form had
part or lot in it ; it sprang from the spiritual need, the love and
the joy of young communities which grew up on all sides, in the
North no less than in the South, to the East of the Holy Land no
less than to the West of it. The several arts of these communities
were so blended as to form a new religious style. To all those
who reject the old superstition of a Roman art one and indivisible,
the first four centuries present a spectacle of the utmost variety
and independence ; forms and types developed side by side in
complete freedom. Uniformity was impossible until communities
merged into Churches and individual States brought Churches
under a single organization, as happened first in Osrhoene, next
in Armenia, and later in Rome and Byzantium. Until the con-
version of Rome, the Christian community in Persia, the second
empire in the then world, maintained an independent existence.
It is clear from the above how necessary it has become for
research to put Rome for the time being in the background.
We are not really qualified to resume the study of Christian art
at the old starting-point until we have made the round of all the
countries important to the problem of its origin. Not till we
have done this are we in a position to follow the clues which by
good fortune the centuries have left in Rome, while in the East
and in the North all traces have been swept away by the storms
of barbaric invasion or by the hostility of other creeds.
It is unscientific to pick out any single source, whether it
be found in a community, a Church, a State, or an artistic tendency,
245' B
2 THE NEW HORIZON
and announce that it alone determined the origin of Christian art.
The early expansion of Christianity was no less rapid than that
of Mithraism, Manichaeanism, or of Islam. While the two
former spread towards the West, and the decisive development
of the third took place in the East, Christianity moved in both
directions, a fact which historians of art are not the only persons
to disregard. Christianity in the first years of its growth embraced
a vast territory inhabited by peoples widely differing in culture,
among whom those of the Mediterranean coast-lands formed
a minority, representing, intellectually, perhaps a third of the
whole. The other two-thirds were Semites and Iranians, both
Eastern peoples ; and it will be one object of the present book to
show how much these two achieved for the growth of Christian
art during the thousand years after Constantine. It must be
remembered that at first there was neither East nor West in the
modern sense ; the distinction began with Islam, and with the
subsequent division of the Christian Churches which left the
countries of the Orthodox Greek Church, as it were, stranded
between the Catholic West and the Orient. The resulting triple
partition of the world must not be confused with the threefold
division which formed my point of departure.
It is still a very general belief that at the beginning of
Christian art the decisive voice lay, if not with Rome, at least with
the Roman Empire, or with Hellenism. Those who so think
overlook the fact that the Roman frontiers ran only just on the
far side of Lebanon and Taurus, and that even before the dawn
of a Christian art Hellenism had begun to give way outside
these frontiers even more than within them. Alexander the
Great, uniting the Mediterranean countries with Persia, created
a great South-Aryan system extending as far as India. This
system was completely dissolved in the period between the first
and fourth centuries of our era through the spread of a strong
rival force which had grown up in the ' Semitic wedge ' dividing
the Eastern and Western branches of the Aryans in the South.
Iran and the Mediterranean, above all, Parthia and Rome, stood
once more confronted, like two irreconcilable worlds ; all
intervening territory, the very lands which had given birth to
Christianity, the lands not sovereign but subject, were imprisoned
between the home of Mazdaism upon one side and that of an
official Hellenism on the other. But what they lost in political
influence and power they recovered by the profession of Chris-
THE NEW HORIZON 3
tianity : Syrians, Cappadocians, Armenians, with the populations
of the adjoining Iranian territory to the East, and the Greeks on
the West, together created in these first four centuries a Christian
art of cardinal importance for the development of the Middle
Ages and the succeeding period, down to the building of St. Peter's
by Bramante and of the church of the Gesii by Vignola. The
great Mediterranean cities, Rome at their head, may soon have
created to their own satisfaction a so-called Christian art. But
the lead was quickly taken by the peoples of Hither Asia, wedged,
as we have seen, between Iran and the Roman Empire, whose
artistic forms were instinct with vital force. These nations were
in contact with Christianity at so early a date that in their case
it was possible for Christianity to become a state religion as
early as the third century : thus both Edessa and Armenia took
this step earlier than Rome. This is a decisive fact not only
in political but also in artistic history ; we must not suffer
our attention to be diverted from it because prevalent academic
opinion does not take it into account.
The earliest research in the field of Christian art discovered
its origins in Rome. It is thus easy to understand how even as
late as 1880 the recognized horizon hardly extended beyond
Rome and Italy. This was about the period when the seniors
of the present generation began their work. It was a time in
which our teachers kept us puzzling by the month together
over the origin of the basilica, and we lived trustfully in the
belief that the catacombs — above all, of course, those of Rome —
were the birthplace of Christian art. The eflFect of all this is
still perceptible in the manuals, for example in Sybel's Christliche
Antike. It is true that Russian scholars did draw attention to
Constantinople, and Dobbert, a native of the Baltic states,
represented the same view in Berlin, but on lines then described
as iconographical, which allowed small scope to central artistic
problems. The result was the creation of a vague borderland
between Rome and Byzantium, not unlike that still conceived
to exist between Roman and Western mediaeval art. All the
initiative was awarded to Rome ; the only matter in dispute
was whether the transformation was effected by the decay of
Roman culture from within or by the victory of Gerrnanic
barbarism from without. Every one breathed a sigh of relief as
soon as he contrived to escape from the Dark Ages to the green
island of the Renaissance. Only there, on the firm ground laid
B 2
4 THE NEW HORIZON
by the humanists, did the historian of art feel himself really at
home ; from there only could he steer the true artistic course.
Antique art, Italian personality — only from the interactions of
these two forces, so the dogma ran, could the new perfection
spring to ' life. Vasari was the prophet ; Jacob Burckhardt's
works on the Renaissance were the scriptures thrust into every
student's hand.
It is clear that the horizon of about 1885 is still commonly
accepted to-day. Wherever the history of art was changed from
a field of free competition into a preserve of official bodies, an
oath of allegiance had to be exacted. All the forces of mediaeval-
ism have been exerted to secure the application of a strait jacket.
If I now treat the origin of Christian art from a fresh standpoint,
I do so in avowed antagonism to this stagnation of a faculty
which had unique opportunities for using the comparative
method to lift scholars over the pale of the usual classical education
and teach them without prejudice to patriotism and without
economic hazard the true position of their country in the world.
But this is not the place in which to criticize the official
administration of art, or to oppose the prominence of those
entrusted with the charge of public and private collections ;
my subject is the present condition of research. Since 1885 the
horizon has been so far extended that many now avoid this whole
field of investigation from the belief that its very extent debars
them from honest collaboration. We have therefore first of all
to consider this enlargement of the field.
First there is the geographical expansion. It began with
a migration from Rome to Byzantium ; we contemplated the
growth of Christian art from a new point of vantage. But the
surviving monuments of the new Imperial city are comparatively
late, hardly a single one dating from before the fifth century ;
what we seek above all are remains of Constantine's own time
or that preceding it. So we advanced to the great cities of the
East Mediterranean littoral, and first of all to those pre-eminent
in the Hellenistic period, to Alexandria, Antioch, and Ephesus.
But what remains to us of these cities ? In 190 1 my book Orient
oder Rom set them in the foreground ; Ainaloff published his
Hellenistic Foundations of Byzantine Art about the same time,
making Alexandria his point of departure. It was the pre-
supposition of Wickhoff's Imperial Art, that Alexandria and the
East only came into the front rank from the fourth and fifth
THE NEW HORIZON 5
centuries onwards. Kraus, inverting the procedure, gave Alexan-
dria the lead in the earliest period and Rome only in the fourth
century. As late as 1901 I was still drawing attention to Asia
Minor and Syria, and above all to Jerusalem. In Koptische Kunst,
my catalogue of the Cairo Museum, and in my book Hellenistische
und Koptische Kunst in Alexandria, it was already recognized
that in all save representational art Christian and Mohammedan
Alexandria was only repaying what it had borrowed from the
East long before. I began to go deeper with Kleinasien, ein
Neuland der Kunstgeschichte, and with Mschatta. I showed the
existence of an Early Christian vaulted architecture and of an
eastern stream of influence in decorative art which even in the
centuries before Constantine was penetrating westward by way
of the triangle of North Mesopotamian cities, Amida, Edessa,
and Nisibis. A number of independent essays followed, till in
1 910 the volume Amida resumed the story of vault-construction,
proved the origin of the barrel-vaulted church in Mesopotamia,
and extended our knowledge of the decorations first revealed at
Mshatta. Whoever wishes to understand the results attained and
the extension of the geographical horizon has only to glance at
the literature which now appeared : Diehl's Manuel de I'art
byzantin, Dalton's Byzantine Art and Archaeology, Kaufmann's
Handhuch der christlichen Archdologie, Toesca's Storia delV arte
italiana and WulfF's Altchristliche und Byzantinische Kunst. But
all of these books still remained more or less at the stage of
Orient oder Rom, or Kleinasien, ein Neuland. My later works
were hardly at all discussed ; in Wulflt's case, from the firm
conviction that Early Christian art, as Ainaloff assumed, had its
origin in Alexandria. My book Amida remains unnoticed ; from
more than one side the attempt has been made to cripple its
eff^ect by depreciatory criticism.
But the catchwords ' Alexandria ', * Hellenism,' ' Rome ',
relate only to the western half of the regions concerned with the
origin of Christian art. They do not include Edessa and Armenia ;
they ignore Mesopotamia and Iran, the countries where Maz-
daism was the religion of court and state on just the same footing
as the cult of the Greek gods in the West. People get round the
difficulty by including these countries among the Hellenistic,
and then conveniently forgetting their existence. Ever since
1902 I have repeatedly called attention to the fact that Hellas died
early in the embrace of the East. To-day my outlook is wider.
6 THE NEW HORIZON
Our idea of the regions important for the origin of Christian
art must undergo a drastic change. By the side of * Rome ' and
' Hellenism ', the West and East of the ancient Graeco- Roman
world, we have to introduce new territories, first Iran, the region
beyond the Tigris ; next, the Euphrates Valley, the Semitic centre
between Iran and the Mediterranean. We can no longer make
shift with the loose phrase * The East ', confining it, as is usually
done to-day, to the east Mediterranean littoral, which means
the great cities of Alexandria, Antioch, and Ephesus with the
inland countries in their sphere of influence. This central region
of growing Christian culture had two other regions on its flanks,
one on the west with Rome as capital, one on the east (Persia)
with Ctesiphon. We usually see nothing but Rome, or perhaps
a little of the Greek religious battlefields in the Mediterranean
area, much as we do when considering conditions within the
Christian Church to-day. We wholly overlook the fact that
before Rome adopted Christianity as her official religion, and
before Byzantium confirmed the Greeks in their old pre-eminence
in the East, there had taken place a peaceful expansion of Christ's
teaching which found its main support in quite another quarter.
This was the line followed by the three books published
since 1910, first Amida, then, in 191 6, Altai-Iran und Volker-
wanderung, and, in 191 8, Die Baukunst der Armenier und Europa.
Behind the coast-lands of the Mediterranean rises as the real
originator, the compact empire of the Parthian and the Sassanian,
where at the beginning of Christian art. Christian communities
were more freely tolerated than in Rome. The development of
events in this part of the world was fixed by definite geographical
conditions.
As soon as Syria and Armenia, Anatolia, Mesopotamia and
Southern Persia, and, finally Northern Iran, were united by the
possession of common ideas, physical conditions themselves
compelled the emergence of a new cultural centre. This actually
happened in the Early Christian period. What the Mediterranean
was to the Graeco-Roman world, that the triangle of cities,
Edessa-Nisibis-Amida, became for the nationalities of inner
Hither Asia : Constantine might have been well advised to found
his second capital here instead of on the Golden Horn. The
real Syria and Asia Minor turn from the coast and look to this
centre ; a Hellenistic highway to India and China once traversed
it, and when barred toward the East, carried Oriental influences
THE NEW HORIZON 7
back in the opposite direction. In Christian times the Hellenism
of Persia and India was already driven into the background by
the reactions which will be treated in their turn in the course of
the present volume, yet it was an influence which can no more
be neglected in considering the origin of Christian art than, for
example, Celtic Christianity in the West. In the radiant focus of
Edessa-Nisibis-Amida I seethe second, the Aramaean centre, round
which the Semites gathered, just as the Western Aryans of the South
concentrated about the Mediterranean, or the Eastern Aryans in
Iran and in the Christian outpost of Armenia. If North Mesopo-
tamia had no political capital such as the other regions possessed
(Osrhoene with its chief city Edessa was swept away as swiftly
as Palmyra), this was only because this frontier-land between
Rome and Persia was too fiercely disputed, perhaps also because
neither rival understood its full importance. This Semitic
nucleus has a value differing in character from that of Armenia,
which was disputed in like manner by the same powers. Both
belonged more properly to Persia than to Rome, yet both aban-
doned in favour of Christianity the Mazdaism of Iran which the
Sassanian princes exploited for imperial ends. Thus North
Mesopotamia lost the political importance which was naturally
its due, to become the centre of an Aramaic-Armenian-North-
Iranian world united by the bond of Christianity against the
Persian. These circumstances alone rendered possible the full
conversion of Edessa and Armenia as early as the third century.
Everything justifies the assumption that fixed types in church
building were here developed before the fourth century, a time
when in the Roman Empire Christianity was barely tolerated or
had only just begun to make itself felt. The supersession of the
Greek style of the Mediterranean by the vaulted architecture of
the Syrian and Armenian area is only intelligible on the theory
that this had already reached maturity.
I. Geographical Influences
We can see why these regions prevailed over the Mediterranean
area in art if we cast a glance at the building materials of the
countries now being christianized, and at the construction to
which they led. While Hellenism and Rome were content with
the timbered church, in the East geology and vegetation forced
men to rely upon other materials and methods. These conditions
8 THE NEW HORIZON
introduced two important values afterwards not less highly prized
in the West, one economic, the other religious.
Iran and Mesopotamia, the region comprised in Early
Christian times under the term ' Persia ', built mainly in un-
burned brick. This necessitated vaulted construction. Archae-
ologists who study the ancient East assume, indeed, that the roofs
of Mesopotamian palaces were always timbered. Whether they
are right or wrong, it is certain that Armenia, from which the
necessary timber was brought down the Euphrates and the Tigris,
was deforested at an early date, and at the period with which
we are concerned was itself compelled to adopt vaulting in all
construction upon a large scale. In Mesopotamia, Babylonian and
Assyrian methods survived as long as builders remained true
to the passage-like hall for which the barrel-vault was the one
possible roof. Here originated the barrel-vaulted church with
transverse axis which formed ongLgQJnt of departure for vaulted
construction in the West.
It was otherwise in Iran. There the dome was indigenous.
In my book on Armenia I have endeavoured to show how this
came about. It would seem that in Iran a decisive influence was
exerted by East- Aryan prototypes in wood-construction, such
as we can still trace in the Ukraine, in Kashmir, and in India.
In these countries the builder roofed a square plan by corbelling
with short beams. As soon as this style of construction entered
countries without timber, like Iran and the Armenia of Christian
times, this corbelling was executed perforce in unburned brick
or in rubble-concrete. So began first the polygonal dome, then
the regular dome, with the transition from the square to the
circular plan by means of squinches. For the further develop-
ment of the dome on pendentives, the reader is referred to my
book on the architecture of Armenia. The decisive cause of the
improvement was the early erection of the dome in Armenia on
supporting arches in place of on the four walls. In course of
time the barrel-vault and dome produced a transformation in the
Hellenistic timber-roofed church, first in the Byzantine area,
subsequently in the West, and this provided the key to the whole
course of architectural development down to and including the
Renaissance. It has seldom happened that any style of building
of oriental origin and wide distribution in the East has failed to
penetrate the West. There are, of course, exceptions, among which
I may cite the remarkable broad-naved church of Eastern Syria
THE NEW HORIZON 9
formed by a series of rib-arches one behind the other, and roofed
with stone slabs, a type perhaps adapted from the architecture
of Arabia.
It was a positive misfortune for studies then in their infancy
that the distinguished Marquis de Vogiie, in his exploration of
the Christian East, should have chanced precisely upon Syria,
a region in which Hellenistic influence became predominant
despite the underlying Arab tradition in the local house building.
From such a starting-point the generalization was sure to follow
that this Greek predominance was universal in the East. It is
easy to imagine how painful was the surprise to find a belief which
had grown into a comfortable habit denounced for distorting the
whole perspective in the development of Christian art. Instead of
personally putting the matter to the test, some of my colleagues
were content to approve the meaningless objections raised by
doctrinaires. The predominance of Hellenism had become their
axiom : they still stand to-day where I stood about 1901 ;
Alexandria, Antioch, Byzantium are for them the sole creative
centres.
But since then we have advanced to a more accurate under-
standing of the compromise reached by Greek and Semite ; we
have followed the destinies of Hellenism in the East, and are
to-day in a position to distinguish contributions made in the infancy
of Christian art by the national architectural forms of Mesopotamia
from those of Southern Arabia, Armenia, India, and Iran. Let
me take one example, the vault. Greek architecture never
admitted this form as a constructive feature on an extensive
scale. It appeared on the contrary wherever timber was scarce
and brick building imposed its own forms. Early examples are
seldom preserved except where unburned brick had been super-
seded by the more durable burned brick or stone. People
therefore jump to the conclusion that the Greeks were the first
to transform vaulted construction from popular into monumental
art, and that this construction began in stone. As a matter of
fact it was brick vaulted construction which, in the field of Christian
architecture, superseded the old Hellenic forms. It can be shown
that vaulting had no firm foothold in the Mediterranean area
in the infancy of Christianity but was an oriental importation
abandoned as soon as connexions with the East were interrupted.
This change occurred in the fourth century alike in Rome and
in the great Hellenistic cities. In East Syria the Arab tradition
10 THENEWHORIZON
preserved the rib-arch and stone roofing. But barrel-vaulted
churches were only erected in numbers east of the Euphrates,
and domed churches only east of the Tigris. These are the
regions which appear to have influenced architectural develop-
ment in Armenia and Cappadocia. Prehistoric vaulting I omit,
as lying beyond our present scope.
Even more instructive is the history of the column and
architrave, the pier and the arch, and of their relations to each
other.
In East Syria you can still see how arch and column were
brought into connexion ; how after the foundation of Constanti-
nople capitals of antique type were modified in the quarries of
Proconnesos on the Sea of Marmara to fit them for carrying the
arch, and how they developed on the lines habitual to Oriental
brick construction with its interior enrichment by decorative
lining ; how the round arch was succeeded by other types, the
horse shoe, the pointed, the oviform, to which diverse conditions
had given rise in the East, and how various other changes took
place to which reference will be made below.
II. Historical Influences
The history of art has followed the example of other history
and of philology. Like these, it has hopelessly entangled itself
in Latin Europe when treating of the West, and the Greek
Mediterranean area when concerned with the East ; it has adopted
a classification of periods in Christian art which those who view
things from the standpoint of the great world-religions can only
regard as erroneous and an obstacle to a right understanding of
the facts. Western Europe is treated as already the centre of the
world in Early Christian and Islamic times, a notion fostered both
by Humanism and by the Church, and now bearing its rich crop
of bitter fruit. The educated Celt, German, or Slav is still blind
to the real state of affairs ; otherwise he could not but regard
history from his proper standpoint as a man of the North, and
classify its periods accordingly ; he could not but abandon the
traditional southern classification into Antiquity, Middle Ages,
and Modern Times. If I am to classify as a student of art, then
the only division I shall make will be between the time when the
East triumphed over Hellas, and rose to predominant place,
and the time when the West attempted to win its independence.
THE NEW HORIZON ii
On these lines, the transitional period begins when Semitic
influence transformed Hellenic into Hellenistic culture ; Iran
and the North completed this movement by their orientation
towards a mediaeval culture. Christianity was the climax of
this penetration. The lead no longer lay with the Hellene, but
with the Semite and Eastern Aryan ; in formative art it lay with
vaulted construction and a decorative system opposed to that of
the Greeks. Those who are obsessed by the classical tradition
of the schools dismiss as a dilettante anyone who holds these
views ; they apply to him trite old phrases of St. Nilus : his ideas
are those of a babe and suckling. To-day Europe is in just such
a state of transition as that which prevailed when the Semitic and
West Aryan elements of the new Christian world first stood on
their defence against Northern and East Aryan culture ; now
as then we see an instinctive resistance to a contrary force which
is not understood. It is the attitude of the Italians to northern
or * Gothic ' art, which the North in its innocence proceeded
afterwards to ape. In their youth both Winckelmann and Goethe
gave free course to their northern blood ; their later tincture of
cosmopolitanism came to them only after they had learned to
know the South.
From this point of view I can regard Christian art as only
one of the attempts to create a new spiritual unity by the contact
of the three Aryan worlds, Western, Eastern and Northern, with
Semitism. This unity, so long as it remained independent of
Church and State, or at any rate was not involved in their attempts
after material conquest, gave a powerful stimulus to artistic
imagination through the compromise which it efi^ected between
man and the world.
In so far, reflection on the origin of Christian art may
gradually inspire new and impressive lines of research. It is in
the power of formative art to bring into so clear a light this active
influence of the Aryan North and East upon the overripe Southern
world that a reaction upon other fields of knowledge may reason-
ably be expected. It is in its power above all to show how it
required ages of war between North and South, East and West,
to bring into being against ' antiquity ', Semitic and Greek, the
culture which we half contemptuously call mediaeval. And that
war is far from ended to-day.
We have to think not of any chronological division in the
ordinary sense, but only of a transition by which, after its sur-
12 THE NEW HORIZON
render to the Semites in the age following Alexander, Hellenism,
the fine flower of antiquity, entered a mediaeval phase, through
contact with Iranian (East Aryan) culture. It will be shown that
this took place about the time of the birth of Christ. The
mediaeval phase will last until the North succeeds in overcoming
the Semitic-Roman conception of Church and State. It is for
all of us to consider how much progress has been accomplished
in this direction, and whether there is a prospect that the purely
northern view will prevail.
After the great migrations and the forward movement of the
East, there was a time when the North was actually in a fair way
to impress its genius upon the life of Europe ; this was the time
when the cities of the West reached the supreme point of northern
expression in their ' Gothic ' art. But Rome made an end of
Gothic. She used the interest of the Italians in their ancient
history as a screen for her ambitions ; by unearthing old Semitic
and Graeco-Roman traditions she succeeded once more in
paralysing the North by driving it into the arms first of humanism,
next of the Counter Reformation, and finally of absolute monarchy.
Art ceased to touch the life of the peoples ; even to-day we hardly
know what it may really mean to man.
It is characteristic that the two groups of persons who seek
to elucidate the beginnings of Christian art, the classical archae-
ologists and the newer historians of art, should be unable to agree
among themselves. It has been rightly said that the first group
views things from the point of view of Antiquity, the second from
that of the Middle Ages, so that the interest of the one centres
rather in the decadence of antique forms, that of the other in the
origin of modern ones. The result can only be a one-sided outlook
for both, quite apart from the fact that neither can obtain any
insight into the regions of real significance for our purpose,
regions historically, geographically, and culturally distinct, but
generally mislabelled by inclusion under that vague term ' the
East'. Thus much we have already learned by our consideration
of the historical and regional factors ; our inquiiy into the
decisive cultural factor will carry the matter yet further. A
survey of this third element will shed additional light on the
adjustments between North and South, East and West, in the
province of early Christian church art.
THE NEW HORIZON 13
HI. Social Conditions
With the extension of the horizon in space went another
extension, Uttle noticed, and due to the conviction that the
origin of Christian art can never be understood without
a knowledge of East Aryan and Northern social conditions.
The beginnings of Islam, which was really a great migration from
the South, and the Germanic descent from the North have always
suggested this line of research, which I sought to inaugurate in
Mschatta and Amida. It soon appeared that the two movements
were not disconnected. In other books — Die bildende Kunst des
Ostens (191 6) and Altai-Iran und Volkerwanderung (19 17) — I tried
to prove the possibility of the connexion. If it is still the general
belief that all this has nothing to do with the origin of Christian
art, it will be the business of the present book to bring about
a change of opinion, but the full development of this question
will be found in another work. Die Baukunst der Armenier und
Europa (1918).
Let us begin by a glance at the attempt, first made some ten
years since, to contract the horizon by introducing the idea of
a Christian classical art. The aim here was to represent the
Christian movement as solely dependent upon the Graeco-Roman
element. A somewhat wider connotation than usual was given
to the term classical by the pretence of including in its sphere
those Semitic lands within the borders of which the Aryan culture
of Hellas ripened. But as a matter of fact the inclusion was
ignored. The idea of a Christian classicism as a decisive factor
in Early Christian art had its birth among the classical philologists.
The theory followed upon the generalization of results obtained
from investigating the Hellenistic sources of Christian art ; it
confined itself almost exclusively to what is called representation,
the objective rendering of things seen, especially human beings,
in their recognized natural forms. We are told that the Hellenism
of Alexandria, Antioch, and Asia Minor advanced to Rome, where
it is accessible to us in the surviving monuments of sepulchral art.
Greek art in its last decline assumed a Christian dress. Everything
beyond sepulchral art, first and foremost church architecture,
the most decisive factor of all, is more or less ignored because
it happens not to be Greek in origin ; there is only one exception
to the ban, that purely practical building, the wooden-roofed
Hellenistic basilica, though in its decoration even this is only
14 THE NEW HORIZON
Hellenistic in so far as its columns were collected from older
structures, which thus served the builders as quarries, as they
later served the architects of the Mohammedan mosques. The
vaulted church of the East, however little the West was able to
follow it in construction, set a new fashion in decoration by the
example of its rich linings ; this fashion prevailed even in the
case of the basilica, despite the representational style of the art
reserved for its walls. But the theory of classical Christian art
is guilty of yet further oversights.
The Aramaic mind, from its centres in Edessa and Nisibis,
set to work to provide a Bible-illustration which even the illiterate
and those ignorant of the liturgical language could understand.
Christian art itself has essentially as little to do with this picture-
writing as with Christian classicism. Artistic quality does not
consist in such representation, but in the nature of the spiritual
content, and in the forms designed to transcend objectivity,
supposing the artist himself to have been moved by any such aim.
Thus in place of Christian classicism we get Christian Semitism.
Such an idea as Christian classic art could never have entered
the mind of the true student of ancient art ; it could only have
suggested itself to the classical archaeologist. The true student
would have discovered that Hellenic art ceased the moment
artists began to abandon the symbolic manner, first by exaggerat-
ing truth to Nature , then by doing the exact opposite , ' representing '
without Nature, and constructing by means of the vault. The
art-student could therefore never have brought down the limit
of classical art so late as the time of Justinian ; he would be more
likely to fix it at the time of Alexander, and then assume a transi-
tional period, in which observation of Nature was first accentuated
in the Semitic fashion, then died out in a counter-movement,
while all the time there were coming into the foreground structural
forms alien to Greek tradition but eloquent of the growing
influence of Iran. The change gradually deprived the Greek
temple of its commanding position ; it lived on indeed, even to
our own day, with its fagade and its orders, but only as a survival.
The real sequence of events is as follows. In the period after
Alexander and about the time of the birth of Christ we mark
the beginning of a new age even in Rome, an age which at first
relapsed into the ways of the old Semitic East, then gradually
found its true course in that artistic sense of the Northern
peoples and the pastoral nomads which drew its inspiration from
THE NEW HORIZON
15
Iran. Christianity was able, it is true, to muster in its defence
a last rearguard of Hellenism. But Islam surrendered wholly to the
new influence, for though it found its most perfect expression
in the cities, its expansion was by way of the deserts ; pastoral
nomads carried it across the world. We could never have failed
to understand this truth had we not suffered an exclusive
devotion to classical studies to narrow our minds, had we not
pored ourselves half blind over books, losing sight of the events
which disturbed the system of the Hellenistic and Roman World-
Empires and replaced their policy by the new aspirations of
Christianity and of Islam.
At the time of the Arab conquests Hellenism itself was
classical no longer. Nevertheless the Iranian spirit rose against
it in unconditional opposition. Borne by the migrant peoples
of the North and East, this spirit began by overthrowing Rome,
then allied itself in Constantinople, as previously in Edessa,
Armenia, and Jerusalem, with the Semitic element in Christian
art, and finally achieved its full development in Islam. The
new art which, after its entry into the West, we describe as
mediaeval, was first developed in the period between Alexander
and Mohammed. Its only connexion with classical art was
through a subsidiary branch preserved by chance in the
Catacombs ; in reality it wrestled with Hellenism and gave it
a fall, triumphantly establishing the Iranian element in architec-
ture and decoration, and in representational art the new Semitic
methods. Hellenism could only be kept alive by the efforts of
learned reaction. It says little for European originality that in
the sixteenth century classical art should have won a new lease
of life with the Counter-Reformation and the art of the contem-
porary Courts, and that to-day it should still preponderate in
the education of our youth. . . .
The above review of the condition of research, based on
a consideration of regional, historical, and cultural influences,
brings us to the following general conclusions. Hitherto the
study of Christian art has been almost exclusively in the hands of
archaeologists. Of these, one group, the specifically Christian,
made the Catacombs its starting-point for centuries. Its members
believed that they could solve the problem of origin by the help
of the oldest Roman monuments, the underground mural paintings,
the sarcophagi, and the timber-roofed basilicas. In recent years
a second group, that of the classical archaeologists, has turned
i6 THENEWHORIZON
its attention to Christian monuments, just as classical philologists
have directed theirs to patristic literature. This was the group
which invented the phrase ' Christian classical art ', the watch-
word of its studies, a phrase which reveals perhaps more clearly
than anything else the one-sided character of such research,
despite all efforts to assert the contrary. A third group, not far
removed from the last, attaches significance to Byzantine Hellen-
ism, and thinks to have discovered in this the creative spirit and
guiding principle of development.
For the specialist who applies the comparative method in the
artistic field Christianity rises against a background formed by
the art of other faiths ; it is the last religion established by a personal
Founder before Islam, beginning, like that faith, in a frontier-
region between the two Great Powers, Rome and Persia. He
therefore considers the appearance of Christian art not only in
connexion with Hellas and Rome, but equally in connexion with
the older great religions of the world, Mazdaism and Buddhism,
and with the younger religion of Mohammed. The archaeologists
base their development of Christian art upon readings of texts
coloured to suit their views and upon monuments surviving in
the Mediterranean area. The specialist is convinced by his
method of wide survey and comparison that this procedure
leaves the origin of Christian art dependent on passages in books
or the existence of single monuments, and further that it makes
Graeco-Roman bulk so broadly as to hide beyond recognition
all trace of other influence under the rank growths of an antiquated
humanism.
IV. Christianity in Persia
This false interpretation of events involves an issue of
decisive importance. Christian and classical archaeologists like
to believe that the teaching of Christ spread only to Carthage,
Alexandria, Antioch, and Asia Minor, and, both by sea and land,
to Rome and Gaul ; they overlook the fact that it spread just as
rapidly towards the East. In the Acts of the Apostles, written
in the second century, there is a list of peoples of surprising
comprehensiveness which has much to tell us on this point.
Among the crowd of Jews who streamed to Jerusalem to carry
the new doctrine out into all the world we read of Parthians,
Medes, Elamites, dwellers in Mesopotamia, Judaea, Cappadocia,
Phrygia and Pamphylia, Egypt and Libya about Cyrene — and
THE NEW HORIZON 17
Romans settled in these parts. The Romans close the list ; the
Persians head it : are they really as negligible a factor in the
spread of the new teaching and in the origin of Christian art as
the accepted theory takes for granted ?
At the beginning of the Christian era Mazdaism dominated
the religious belief of the East in just the same way as Hellenism
dominated that of the West. The manner in which the cult of
Mithras spread throughput the Roman Empire and that of Anahita
over the whole of Hither Asia proved that it must be reckoned
among the world-religions. Armenia was Mazdean ; in Persia
proper the doctrine of Zarathustra was exploited for political
ends after the Semitic fashion by the Sassanian dynasty, which
made it the State religion to prove their national monarchy
established by the grace of God. Are we to suppose the immense
religious force of the Zend Avesta, exerting through Persia and
Armenia so widespread an influence towards the West, a negligible
factor in comparison with Hellenism at the birth of Christian art ?
Opponents will meet this question with another : where is the
evidence for a Mazdean art ? I will refrain from citing the
Persian rock-reliefs, where Ahuramazda is represented either
mounted or on foot. Another question is really of more signifi-
cance : is it so certain as is commonly supposed that Mazdaism
had no widely distributed and popular art ?
The idea that a whole religion can have renounced art is
inadmissible, and the facts are against it. The aims of religion
and art are too ahke for the one to dispense with the other ; both
seek to establish and to reveal an invisible inward world of human
hope beside the actual and visible world around them. I say
nothing of prehistoric beliefs, or those of existing primitive
peoples, though these too seem hardly ever able to dispense with
symbols which may be regarded as the beginnings of formative
art. In this book I am dealing only with the two main types of
historical religion : that which emerged from dim beginnings
in the forcing-houses of southern culture, and that revealed by
personal founders, issuing more or less from the beliefs of northern
peoples and pastoral nomads, and wholly or in part displacing the
first. To say nothing of Moses, it is no mere chance that Buddha
and Zarathustra, both Aryans, stand at the head of the list.
Works of formative art are among the most important
sources of evidence for all religions. The Egyptian spirit is
essentially opposed to the Hellenic, but both achieve their
i8 THE NEW HORIZON
highest in their works of art ; every one knows what the monu-
ments left by Asoka did for the expansion of Buddhism, and what
those left by Constantine did for that of Christianity. Not long
ago we learned for the first time from the funeral monuments of
the Han period the nature of the state worship established by
Confucius ; in the ' philosopher landscapes ' of the Sung period
the critic detects the influence of Taoism, perhaps also the trace
of Indian thought. The art of all these cults represents ; and
we are so used to finding their religious ideas in their artistic
creations that Islam, the only familiar world-reHgion which does
not represent, is left out of the account, as something abnormal
in the religious sphere and to be neglected with impunity.
In my book Altai-Iran und Vdlkerwanderung I attempted to
show that in the perfected style of Islam there still survives that
non-representational northern and nomadic art known to us
through the work of prehistoric times and that of the later
Teutonic and Turkish tribes. In the period of the great migra-
tions, both these races advanced towards the ancient forcing-
houses of culture, just as the Greeks, Celts, Persians, and Indians
had done in pre-Christian times. Originally none of these peoples
represented ; they first learned this mode of artistic expression
in the South. The student of art is inclined to think that the
contrast between North and South may be explained by the
transition to a higher stage of culture. In the South, man passed
immediately from the culture of the earlier Stone Age into a social
system which sought to cast a spell upon the object by repre-
sentation, as the primitive hunter attempted to do when he
made pictures of his game. In the North, on the other hand,
formative art developed out of the handicraft of the later Stone
Age. It enclosed space in borders and filled it with ornament
which for the most part followed from' the nature of the material
and the process adopted, ornament which was therefore geo-
metrically designed for the purpose of pleasing the eye.
Judging from the ornament of Ostrogothic and Lombard
metal work and sculpture in Italy, from Irish illumination and
from northern antiquities generally, but above all from the oldest
Scandinavian wooden churches, the student of art cannot but
infer that the North hardly knew Christian ' representation ' at
all. For the moment we may leave in the background its pre-
Christian art, of which the roots are to be found in Celtic and
Germanic religion.
THE NEW HORIZON 19
To this northern group, or to that of the pastoral nomads,
belongs Mazdaism, the only world-religion bom in the sixth
century before Christ on the confines of the North, but destroyed
by the united advance of southern religions. Buddhism,
Christianity, and Islam. Like these, it owed its origin to a
personal founder ; its scriptures, burned by Alexander, but later
collected in the Zend Avesta, were in use among the Iranian
peoples for more than a thousand years. Neither the Indian
sacred books nor the Bible could make head against them ; the
Koran alone succeeded in displacing the Avesta after the Persian
reHgious spirit, represented by the cult of Mithras and by
Manichaeism, had proved itself a dangerous rival to Christianity
even in the West. It is strange, but no one seems to have reflected
that like all other religions, and especially those originating in
the South, Mazdaism must have had an art. The circumstance
that none of its monuments have been preserved or discovered,
at any rate in wide distribution, no more proves that such an
art did not exist than the similar lack of documentary evidence.
A little while ago, whoever dreamed of the monuments of all
kinds recently discovered in Chinese Turkestan ? Yet there
they are, and in such overwhelming abundance as to make it
incredible that we should have remained so long without the
faintest idea that so brilliant an art existed. For this ignorance
we have to thank the perverse method of our research, which
only accepts as worth notice, or admissible for scientific work,
what we can see embodied in surviving monuments. Since
Winckelmann's day, everything possible has been done to
rediscover the antique ; but in the meanwhile the other gaps in
our knowledge have been ignored. Such a gap, and a wide one,
hides from us the creative activity of Iranian Mazdaism in the
first period of Christian art. I have therefore made it my special
task to investigate it and to ask the following question : has
Christian art really no constituent features allowmg us to pre-
suppose the existence of a Mazdean art ? The argument by
presupposition will be found running like a red thread through
the following chapters.
While still fresh from the impression of a new world opened
to us, let us now consider a pronouncement like the following ;
' Classical antiquity finishes its course in Christian art, and
* accomplishes its destiny. As far as painting and sculpture are
* concerned, the art of the imperial age, including the Christian,
c 2
20 THE NEW HORIZON
' moves on a descending line ; there is no suggestion of new
' development ; we find no well-spring of youthful energy, nothing
* but the decay of hoary age. Beauty survives only in the earliest
' productions of the period. For the rest, the significance of all
' this work is not to be sought in its aesthetic quality, but only
* in its subject matter, in the origin of a Christian iconography
* and in its value as a key to early Christian thought. Yet in
* architecture, antiquity, and precisely Christian antiquity, was
* creative during the later imperial period ; it must be admitted
* that here it was still able to celebrate a final triumph.' Such
monstrous perversions of fact — I speak from the point of view of
the student of art — can only come from scholars buried in the
dust of antique learning. For what are the facts ? Long before
Alexander there were great national states in * the East ', in
Egypt and in Mesopotamia. These gradually lost their national
character, and Egypt was eliminated as a creative force. Not so
Asia. The classical archaeologist fails to see that what made
Greece great, her Aryan quality, was first felt in its full vigour
at the birth of the two new world-religions, Christianity and
Islam, and not in the Semitic area, but beyond it in Iran (I leave
India for the moment out of the discussion). With these religions
it matured its powers, with these it flourished, and by their help
ended by achieving the artistic conquest of East and West alike.
The problem to-day no longer concerns the rival claims of
Rome and the East ; it concerns the Aryan spirit in East and West,
its assimilation of religions which had grown up on Semitic
ground, its flowering through their formative art. Just as
Christianity took final shape among the western Aryans, so did
Islam amon^ the Aryans of the East. The subject of the present
book is limited to the period between Christ and Mohammed.
But in this period Christian art itself was penetrated by the
East- Aryan spirit ; its whole development towards its approaching
ascendancy in Europe is only intelligible if we grasp this hitherto
neglected fact. Early Christian art was only antique, that is, Greek
and Semitic, in those features which crippled its development, in
the timber-roofed basilica, and in the monotonous objectivity
of its representation. Architecture and its decoration, which gave
art life and growth for more than a thousand years, which formed
the germ of its development, was entirely in the hands of the
East ; those northern Aryans who had come into contact with
the South understood the Oriental forms, adopted them, and
THE NEW HORIZON 21
perfected them in freedom from the oppressive influence of
Semitic monarchism. These are the true facts about the develop-
ment of Christian art : in the Hght which they throw, let any
man decide whether the creative force rose from the well-springs
of youth or from the places of senile decay. What Hellas was to
the art of antiquity, that Iran was to the art of the new Christian
world and to that of Islam. So at a later time the northern spirit
informed the art which we call Gothic.
It must be remarked of the history of research in the field of
Christian art, that those who pursue it, whether they reject or
approve, have always kept within the limits which 1 have pro-
gressively defined ; in following the plan of my life's work, they
have never advanced beyond me and so proved their compre-
hension of my task as a whole. They have been drawn on step
by step, first by my work on Constantinople, then by that on
Egypt, Asia Minor, and Syria. At the present moment we are
crossing swords over Mesopotamia; to-morrow it will be over
Armenia, the day after to-morrow, over Iran. Each time they
will begin by resistance and proceed by invading my claim and
trying to drive me out. The day may come, I may yet live to
see it, when they will envisage my course as a whole and at last
begin to join their forces with mine instead of blocking my path.
Speaking generally, we have at present advanced no farther than
agreement as to the significance of the Hellenistic cities on the east
shores of the Mediterranean ; first Alexandria, then Antioch and
Constantinople with their closer relations to Asia Minor ; in short
they accept my point of view down to the publication of Kleinasien
in 1904. On the other hand they still reject the high antiquity of
vaulted construction in inner Asia Minor and Mesopotamia, ascribe
the initiative here also to Antioch, and even give Alexandria the
credit for features of a pronounced Iranian character ; they there-
fore question the existence of any vaulted buildings before the
seventh century, and explain them by hypothetical later recon-
struction. I trust that the time has now arrived when instead of
contenting themselves with manuals uncertain of their ground,
people will rather listen to the specialist who has given thirty-five
years of his life to the problems of development in this field of
research. The bibliography at the end of this volume may incite
my readers to adopt this course.
The church-building of Christian Persia had two distinct
characteristics : it employed the vault from the very beginning ;
22 THE NEW HORIZON
it decorated walls with linings. The vaulting may be either
domical or of the barrel variety. Both kinds seem to have existed
independently when Christian church-building began, but not
to have been used together until the Church in the Mediter-
ranean area demanded a building with a longitudinal axis. I am
disposed to ascribe the origin of the barrel vault more especially
to Persian Mesopotamia, that of the dome to Iran. The distribu-
tion of vaulted churches supports this view, a matter to which
I will return later.
What vaulting is to structure, that the lining of the vaults
and walls is to pictorial art. Properly speaking Iran knew nothing
of the graphic arts in the narrower sense, since it did not represent,
but confined itself to pure decoration. The Iranian style is thus
at once and fundamentally distinguished from that of ancient
Mesopotamia, which covered walls with representation. We all
know the wonderful naturalistic reliefs in alabaster from Nineveh
now in London, and the bodyguards in glazed earthenware from
Susa in the Louvre. But in addition to such representations of
natural forms, Mesopotamia had splendid ceilings with repeat-
patterns enclosed in borders. Was this also purely Semitic ? If
m Mesopotamia we find representation and purely decorative
ornament side by side, we must insist upon the fact that originally
the ornament occurred alone in north-eastern Iran, and marks
a stream of influence wholly Aryan. Even the Greeks when they
migrated into Hellas, and before they came into closer touch with
southern art, had, it would appear, only a geometrical and non-
representational art of this nature.
It has been generally deduced from the study of Islamic art
that its manifold designs and patterns were accumulated by
voluntary contributions from all parts of the Hellenistic world
and attest the penetration of monumental art by the various
industrial crafts. Those who offer such an explanation reveal
a profound ignorance of Iran. It may be said in their excuse
that in their treatment of Persian art they set out from the
monumental art of the Achaemenian and Sassanian South, where
first the Semitic element, and afterwards the Hellenistic, drove back
and overgrew the Aryan. But after all it is still quite possible
to infer the nature of this Aryan element from buildmgs in which
various kinds of wall-covering have been reproduced in stone,
and the evidence of their existence thus preserved. Such monu-
ments are Mshatta, part of the fa9ade of the great mosque at
THE NEW HORIZON 23
Diarbekr, and the Stupa at Sarnath in India. It is true enough
that Iran itself has so far yielded very slight traces of the pre-
Mohammedan period ; but this is explained by the fact that the
building material was unburned brick. When this fell into ruin,
the lining of the walls fell with it. We have also to remember
that systematic excavation with an eye to this kind of discovery
has yet to be undertaken. When our Austrian expedition wished
to excavate at Naishapur, the Franco-Persian agreement was put
in force to stop our work. But the excavations at Samarra have
confirmed the theory which I first conceived in Egypt and later
developed in my books Mschatta and Amida ; that the moving
force in the art of Islam came immediately from northern Mesopo-
tamia but ultimately from the more distant centres of Iran.
V. The Asiatic East
In conclusion I will add a few words on the historical
expansion of Christianity towards the East, first into the Persian
Empire, then beyond it into India and China. In the West the
new faith had only stopped with Ireland and the Atlantic. Its
expansion in the East was just as wide : Chinese Turkestan is
about as far as Ireland from the Holy Land. Here the exploration
of the last two decades has yielded fifth-century fragments of
a Psalter in Pehlevi, and of another Psalter in a Syro-Persian
dialect, while Sogdian MSS. afford a glimpse of early liturgy.
This variety of races thus represented among Christians in these
remote regions may perhaps serve to explain why it was that
Christianity was never politically as creative as Indian Buddhism,
or the immigrant Manichaeism which entered Central Asia by the
same Persian route. In the eighth century, indeed, tHe Nestorians
did very nearly bring about a change in this state of affairs. In
China a decree of a.d. 638 proclaims their arrival with sacred
books and pictures, and the erection of a church in the capital
Si-ngan-fu which in the eleventh century was known to the
Chinese as a ' foreign Persian temple '. The priests came * from
the gold-bearing districts ', perhaps Tokharestan and Bactria, of
which the capital, Balkh, was then a great centre of Nestorianism.
The emperor Kao-Tsung (a.d. 651-84) caused Christian churches
to be erected in all the ten provinces of China : some of these
were restored in wood and stone by Hiian-Tsung (a.d. 713-56),
who also caused services to be held in the Hing-King palace.
24 THENEWHORIZON
Su-tsung (a.d. 756-63) built churches in Ling-wu and four other
places ; Tshi, a priest from Raja Griha in India, who had risen
to high office at Court, once more restored the old churches,
added to their number, and gave them so rich a decoration ' that
they resembled the plumage of the pheasant in his flight '. In
his time (a.d. 781) was erected the stele in the capital, in the inscrip-
tions on which all these facts are recorded in Syriac and Chinese :
' Great palaces of light and unity (churches) covered the length
' and breadth of the Middle Kingdom.'
This East-Iranian Christianity entering China through the
country beyond the Oxus and through Turkestan, and found
flourishing there in the seventh century, started from the
district of Edessa and Nisibis on the middle course of the Tigris
between Diarbekr and Mosul. We learn much of this Parthian
region from the Acta of St. Thomas. The above Chinese
description, ' a foreign Persian temple,' has already given us
a clue to the style of the churches. More precise is a remarkable
passage in the Bar Saba legend of Merv.^ Here the first church
in that place is said to have been built on the plan of the Parthian
Ealace at Ctesiphon, traces of which would thus appear to have
een still visible in the Middle Ages. It may be assumed that the
Parthian palace type is represented by the two surviving examples
of Firuz Abad and Sarvistan in the province of Persis, where the
central part of the building consists of one or more square halls
covered by domes. We shall find the earliest churches in
Armenia immediately connected with halls of this type ; for the
moment, however, we are only concerned with the antiquity of
church buildings in Iran.
Along the east bank of the Tigris extended a province of
which the early Christianization and the art are so well attested
by historical evidence that its neglect by workers in our field is
a matter for some surprise : this district was known as Adiabene,
with its capital at Arbela. The most important document for
the early Christianity of this region is the Chronicle oj Arhela}
The writer is Mesihazekha, a pupil at Nisibena in the middle of
the sixth century when Abraham (a.d. 509-69) presided over this
school of Eastern Christianity. His chief source was the teacher
Abel, important to the historian of art as one of the witnesses to
Christianity in the Parthian period , that is , before A . d . 226 . Which
^ Communicated by E. Sachau. ^ Edited by E. Sachau.
THE NEW HORIZON 25
of us ever expected to learn of a church architecture flourishing
beyond the Euphrates and the Tigris as early as the second
century of our era ? My own researches in Armenia had indeed
already forced me to this conclusion, but the Chronicle of Arbela
brought direct proof. Abel the teacher, living about a.d. 200,
may well have derived information from the episcopal archives
of Arbela ; other facts he doubtless obtained from local tradition.
There were churches in the place the builders of which must
still have been remembered. The church of Isaac standing ' to
this day ', that is to say the time of Mesihazekha, was built by
Isaac, third bishop of the city (a.d. 123-36), and restored by
Abbusta the eighteenth bishop (a.d. 450-99). We have no exact
knowledge of its type ; the church was large and well propor-
tioned ; when it was restored, it was enriched ' with all possible
embellishment '. In memory of Noah, fifth bishop {c. a.d. 166-71),
a second church, known as the small church, was erected ; ' its
site was still known,' though by the middle of the sixth century
it was itself no longer in existence. Its builder was perhaps the
teacher Abel, if this personage is to be identified with Abel the
sixth bishop (a.d. 171-c. a.d. 200) ; and within its walls the ninth
bishop, Sahlupha (a.d. 235-41) was interred.
Facts like these give us some idea of the strength and wide
\- distribution of Christianity in Persia, both Iran proper and
Mesopotamia, enabling it to exert an influence far into Eastern
Asia beyond the intervening region of the Altai. Nestorianism,
which we found at work in China, was of relatively late appearance;
it began with the opposition of the Aramaic populations between
the Euphrates and the Tigris to the decisions of the Council of
Ephesus in a.d. 431. Even before this, in a.d. 410, an episcopal
constitution had been framed in Mesopotamia with the CathoUcs
of Seleucia at its head, of which the sphere of operation extended
beyond Media and Parthia into the regions whence the first
priests entered China.
Christianity was transplanted into Southern Iran from
Antioch by an exercise of arbitrary power. The eleventh-
century Chronicle of Soort, with which Graeco-Roman sources
are in essential agreement, records that after his conquest of
Nisibis and Antioch in a.d. 256 and 260, Shapur transported
prisoners to Babylon, Susiana, and Persis, where he gave them
lands and houses. In consequence, Christians became numerous
in the Persian monarchy, and monasteries and churches were
26 THENEWHORIZON
erected. In later sections of this book we shall have to notice
the traces left by this Antiochene movement as it ebbed back
from the south of the Sassanian territory.
It was from the upper course of the Euphrates and Tigris
that the influence came which sought to dominate the National
Church of Armenia in the fifth century. In that country the
ruling Arsacid dynasty had established, about a.d. 300, a State
Church which was architecturally dependent on East Iran, and
derived its objection to representational art from the Mazdaism
hitherto prevalent in the land. When the Armenians invented
an alphabet of their own and set about the creation of an ecclesias-
tical literature by translating Syrian and Greek books, close
relations began between Armenian theologians and those of
Nisibis and Constantinople. The effects of this are marked in
the province of church architecture by the efforts of the Armenian
bishops to replace the national type of church, with central dome
and radiating limbs, by the barrel-vaulted church with long nave,
and to substitute North Mesopotamian or Greek wall-paintings
for the old non-representational style of decoration. Only the
abandonment of Chalcedonism ^ gave their opportunity to the
opponents of these innovations, who were not only strong in
popular support, but were also represented, if somewhat uncer-
tainly, in the Armenian Church itself. For though the Parthian
Arsacids died out, driven from the throne in a.d. 428, the
architects of the Nacharars were still able to impose the long nave
on the national form of church (see below, p. 68). In Armenia,
at least, we clearly mark the triumph of the national style of the
fourth century over the ecclesiastical influences of the fifth, after
a struggle lasting for at least a century. A similar observation
may be made in the case of East Syria ; there too the national
Arabian mode of building was pressed back in the fifth century,
only to be more strongly reasserted at a later time. But in Syria
Christian architecture came to an end as early as the beginning
of the seventh century with the crippling of the Byzantine
Empire, and was finally displaced by that of Islam. Asia Minor,
on the other hand, remained Christian for another five hundred
years ; while in Armenia, from the ninth to the eleventh century,
the arts reached a high development, this kingdom remaining
permanently Christian in despite of Mohammedan suzerainty.
^ The Council of Chalcedon held in a.d. 451 condemned the monophysite
heresy.
THE NEW HORIZON 27
These were the regions from which domed architecture passed
to Constantinople, to the Greek Church, to the Balkans, and to
Russia, while Mesopotamia transmitted the barrel-vaulted type
of church with long nave to the Latin Church in the West.
Brunelleschi once more recovered the old tradition of domical
building, which in the hands of Leonardo, Bramante, and Vignola
became the predominant style in the Europe of the late
Renaissance.
II
Community, Church, and Court
THE revelation by the artist of the beliefs determining
his personal outlook on the world should form an
essential part of every religious movement in the field
of art. Such revelation there was in the case of a Leonardo,
a Michelangelo, a Giorgione, a Diirer, a Rembrandt ; but even
by them it was only made at times when they were creating
independently and following their own nature. As a general rule
the patron steps in between the artist and religion, sometimes in
the shape of an individual, at others of a corporation, above all,
naturally, in the shape of the Chtirch ; the patron intervenes
either in a general way through an official right to prescribe and
judge, or in his individual capacity, himself giving the commission
for the work. The result of these relations is a material restriction
of the artist's genius which must not be left out of account when
we consider the origin and early growth of Christian art. Can
a religious art allowing full personal freedom be said to have
existed at all in the early Christian period, or was all artistic
creation from the very first under such constraint ?
A general tutelage of the earliest Christian art on the part of
the Church or the Court is incredible. Before anything of the kind
could have occurred, a Church would have had first to be organized ;
it would then have had to win so great a power over the masses
that even the creative artist would have been forced to obey
its authority. But the outstanding feature of the earliest Christian
art was its purely religious development among communities
distributed over different nations. This aspect of the question
I find imperfectly considered in the literature of Christian
archaeolo^. It must, however, become the subject of scientific
investigation, since it is more than possible that art, at first
isolated by differences of nationality and geographical position,
developed under the impulse of spontaneous aspiration. The
COMMUNITY, CHURCH, AND COURT 29
Church in the meantime increased in power, but was seldom
creative. Its course was to retain such artistic elements as accorded
with its aims, and to reject everything which did not.
The first of these two kinds of growth presupposes a high
level of moral and intellectual life ; it presupposes an eagerness
to solve the riddle of existence, and artists with ideals beyond the
aims of every day. But does not the very creation of religions
by personal Founders itself involve the existence of this high
level ? Religions, even that of the Greeks, which grow gradually
from obscure beginnings, only attain the power of artistic expres-
sion by slow degrees ; on the other hand, religions established
by Founders may at the very outset reach the supreme point of
their artistic achievement. Christianity is a religion of this latter
kind ; about a. d. 400 its foundations were already completed.
But here we must carefully distinguish between the East
and the West. In the West it is probably quite true that the
rigid organization of the Roman Empire suppressed all national
movements. In the East, on the other hand, this central organiza-,
tion was not effective beyond the coast-lands of the Mediterranean.
In the interior lay the different ethnical groups represented by
Cappadocians, Armenians, Syrians, Jews, Egyptians, and beyond
all these the second great power of the age, Parthian and Sas-
sanian Persia, with an official religion, Mazdaism, more tolerant
than the State Hellenism of the Romans, and admitting a freer
expansion of Christianity than any sanctioned in the West.
The case of Armenia shows us how these conditions reacted
on the peoples wedged in, as it were, between East and West.
Here an Eastern Christianity overthrew Mazdaism and was
accepted as the religion of the State before any such change
was contemplated by Rome.
Thus even in late Hellenistic times the Christian art of the
East was rich in features which could only have grown up
naturally in countries living their own national lives. In archi-
tecture, such a feature was the use of vaulted construction,
resulting in the principal types discussed in the next chapter.
In decoration there were the contrasting systems of representation
and abstract ornament, and, within the limits of the first, the
symbolic and the realistic styles, to be discussed in Chapters V
to VII. All these groups enjoyed their independent existence
during the first three centuries ; in the East the independence
continued during the fourth. In the fifth century the ecclesiastical
30 COMMUNITY, CHURCH, AND COURT
organization, now established in the Greek and Aramaic areas,
attempted to subject the local and popular influences ; in the
sixth century the secular power, in alliance with the Church,
followed the same course, commanding the vastly greater resources
of imperial Constantinople. In the East the beginning of deca-
dence coincides with the appearance of Islam. On the outbreak
of iconoclasm, the forces which had been developing an art
averse from Hellenism shook the foundations laid by Church
and Court. This led to a fresh division of the artistic world.
Great individual developments ensued both in the West and in
the Mohammedan East, while Constantinople, placed between
the two, ate the stale bread of old traditions. This view of
evolution in Christian art will be established in subsequent
pages ; here I must lay stress on the fact that it is based solely on
many years of experience in the study of the actual monuments.
None, therefore, should be surprised to find things wearing
an aspect very different from that which they used to present
in the exclusive light of literary evidence.
I. Christian National Art dozen to A. D. 400
Just as there was never a single and uniform Christianity, so
from the very first there was never a single stem, far less a smgle
root, from which Christian art sprang. On the contrary, the time
when there was still no uniformity in the Church and still no
Christian State was precisely that when the controlling factor
was the national spirit, which varied from one people to another.
According to the received idea, Christian art rose as a homo-
geneous growth from Hellenistic and imperial Roman ground.
We have seen that these limits must be extended. The moment
the geographical horizon is widened, the vital and creative force
is discovered in local genius. Race, nationality, and economic
condition count beyond question for more than political and
intellectual connexions. Historians of art seem to have lost the
power of reckoning with these essentials ; with their narrow
European standpoint, their purely classical and philological
outlook, they hold it rather unscientific than otherwise to take any
cognizance at all of these important matters. My own belief
is that without a fundamental study of these things an answer
to the problem of the origin of Christian art cannot and should
not be attempted.
COMMUNITY, CHURCH, AND COURT 31
Christ appeared in Judaea, under religious conditions already
determined when the Jews were still a wandering people. For
the study of art, the important fact about the Jews is not that they
were Semites, but that they never created an art of their own.
The case of Judaism resembles that of Islam. Mohammed
appeared among Bedouins, nomad Arab herdsmen, who had no
more art than the Jews ; in any case neither people possessed
that representational kind of art which we have come to regard
as characteristic of Jewish life.
Semitic art, as a species, flourished in river valleys and
among their settled agricultural populations. These peoples
created Church organizations and States with their corres-
ponding artistic tendencies so deeply rooted and so indigeneous
as to ensure a regular and progressive development. Conditions
were different for nomad herdsmen, as I showed in my book
Altai-Iran ; I shall return to the subject in the fifth chapter of
the present volume.
When in 1901 I put the question : The East or Rome ?
and later added the second question : The East or Byzantium ?
I was thinking in the first instance of these Semitic or Hamitic
peoples who determined the course of ancient history in Mesopo-
tamia and in the valley of the Nile, peoples whose national sense
increased in strength m Christian times and influenced the world
through the great cities of the eastern Mediterranean. More
than ten years work on Mshatta, Amida, Altai-Iran, and Armenia
fundamentally changed my point of view. I still hold that at
the beginning of Christian art the Semites played a decisive part.
Most important among them at this time were the Aramaeans of
the northern parts of Syria and Mesopotamia, the Nabataeans
forming a link which connected them with the Copts. Hitherto,
when we have spoken of the Christian East, we have really meant
these Semites of the interior behind the Hellenistic coast-lands.
The time has come to modify the once suggestive formula ' the
East or Rome ', in so far as the growth of our knowledge is now
leading to the analysis of the term ' East '. In making any
change I have in view not so much local as racial division.
Behind the Semitic world was a world of Aryans. In Christian
times Eastern Iran was inhabited by a great population not to be
measured by the same standards as the Semitic peoples. Perhaps
we shall best understand the true nature of the Aryans of Iran
if we compare their art with that of the Hindus. These Aryan
32 COMMUNITY, CHURCH, AND COURT
immigrants into India knew no representation ; they adopted it
from the natives of the country, just as the Mediterranean
Greeks had adopted it from the Semites. The Eastern Aryans
knew nothing of representation. The Achaemenian, and later the
Sassanian, court adopted the Semitic representational manner ;
but the people, though their art had been much enriched by
pictorial motives, yet remained true to their surface-filling
ornament. Now this loyalty to Aryan and northern feeling
had a part to play at the beginning of Christian art. If the
Iranian point of view had already influenced the Semites and
the Greeks, the origin and early growth of Christian art become
unintelligible without it. As Islam issued from a southern
people with no representational art, quickly finding its main
support in Iran and amon^ the nomad herdsmen of the Iranian
North-East, so Christianity issued from the Jews, a nation
in like case, and found before long its strongest support among
Persians and Armenians, both East-Aryan peoples. In the
case of Islam, the bond of union was always the Arabic tongue.
In that of Christianity, it is true that in the Mediterranean
area the Greek and Latin languages formed the bond of
union. But in the East there were national groups wholly
independent of the Mediterranean peoples, who translated the
scriptures of the Greek Christians into their own tongues —
Aramaic, Armenian, and Pehlevi. In like manner artistic forms
were exchanged. My work on the architecture of the Armenians
and its influence on Europe initiated research with regard to
one only of the above three national groups ; the present book
will deal more comprehensively with the subject. For the
moment I am concerned with the racial question in a general
manner.
When Christian art came into being, the Jews were no more
independent than other Semites ; they were completely subject
to Aryan peoples — in Mesopotamia to the Persians, in Syria and
Egypt to the Romans. The fact leads us to observe that even
at this period we have to draw a sharp dividing line between
Eastern and Western Aryans. From the separation of the two
sections by the Semites, it resulted that they could only have
a common frontier where the Semitic wedge came to an end, on
the upper course of the Euphrates and Tigris, that is, in Armenia
and, to a certain extent, in the North- Mesopotamian triangle of
cities : Edessa, Nisibis, Amida. Was it mere chance that the
COMMUNITY, CHURCH, AND COURT 33
first Christian states came into existence precisely at these points
of contact, in Osrhoene and Armenia ? Be this as it may, the fact
had momentous consequences for art.
In A.D. 323 Rome raised Christianity to the dignity of the
state religion ; but when soon afterwards Constantine founded
his new capital on the Bosporus, he arrived too late for his
action to affect art, just as before him Rome had come too late
to meet Greece as a rival in the intellectual and artistic fields.
Christian art was already in growth or full existence, and * Chris-
tian classical art ', which means West- Aryan art, was helpless
to prevent the triumph of the East. The two factors which
made this a certainty were in the first place the presence of the
great Semitic centre of Aramaean North Mesopotamia ; in the
second, the support of the Armenian people, the extreme outpost
of East- Aryan culture.
Thus in the search for the origin of Christian art we have
to divide the surviving monuments into three groups. First,
that of the Mediterranean area, which itself falls into two sub-
divisions, a Greek and a Latin. Egypt belongs to this group
only through Alexandria and its immediate sphere of influence ;
Syria only through her coast-lands, with Antioch as their
intellectual centre ; Anatolia, in like manner, only through the
littoral and through its western half. There is m the second
place the intermediate Aramaean region, with its unrivalled
intellectual centre in Edessa-Nisibis ; it includes Mesopotamia
and the interior of Asia Minor, and comes into touch with the
real Egypt through inland Syria. In the third place there is
Iran, revealed to us through literary sources as extending its
influence to East Asia, but so far as surviving church-buildings
are concerned, as yet only known to us through its western
offshoot, Armenia.
The threads first run together in the Hellenistic centres on
the Mediterranean littoral, next in Jerusalem and Constantinople,
when the two cities were re-founded in the time of Constantme.
Further, the compromise between the West-Aryans of Asia
Minor with the East-Aryans of Armenia and the Semites of
North Mesopotamia is a most notable factor in the origin of
Christian art. There may still be talk of a ' Christian classical
art ' of the Mediterranean, culminating in Alexandria, Rome,
and the maritime cities of Asia Minor. But it will no longer be
permissible to ignore the existence of a Christian Iran, and of
»45» D
34 COMMUNITY, CHURCH, AND COURT
Semitic Christians who made their influence felt in art after
a fashion which had been quite impossible to the Jews. Such,
more or less, were the racial and national foundations of Christian
art in the first three centuries, and even in the fourth, Rome for
the first time brought Christianity into imperial politics.
The artist working in a country having an artistic sentiment
of its own is able to provide a free religious community with
forms corresponding to its needs, for instance, with an acceptable
type of church ; he is in harmony with national feeling and ideas,
and within the limits which they impose is free to obey his
unconscious impulse. No autocratic patronage compels him ;
the result is that his work develops a local character, stronger
and more durable than any general bond, political, linguistic,
or intellectual, imposed from above, as in the case of the Hellenistic
or the Roman system. Even when new Christian Powers arose,
like the Byzantine Empire, it was long before the assertion of
their strength made any serious impression on the different
nationalities of the East. In my own belief the whole fourth
century still belongs to the epoch of individual national move-
ments, at any rate in the province of art. As we shall see below,
two occurrences give a clear indication of attempted ecclesiastical
control : the penetration by the long church of regions originally
unfamiliar with it, and the dissemination of figure art through lands
unacquainted with representation. It was, apparently, not until
the fifth century that the long church began to influence the
national style of Armenia and Mesopotamia ; in Eastern Syria,
at any rate in the three-aisled type, it was rather earlier : single-
hailed churches need not be considered, since they may be found
anywhere, with any kind of roof. The intrusive art of repre-
sentation which entered Armenia at the same time, whether
from Edessa-Nisibis or from Constantinople, has no longer the
bright symbolic character which it bore among the Greeks, but
had already assumed the didactic and directive manner of
Semitic art ; of this we shall have more to say in a later chapter.
Here, too, the change was not felt before the fifth century, a point
which admits of demonstration. Thus down to about a.d. 400
the several Christian communities, united in national groups,
had developed Christian art on their own lines ; the fifth century
marked the first attempt to bring these separate units under the
control of a common Church. As in other cases, so in this, my
first insight into the truth came from Armenia. A third move-
COMMUNITY, CHURCH, AND COURT 35
ment in the direction of uniformity, the transformation of
classical naturalistic art into diffused ornament in a single plane,
resulted from purely artistic rather than from ecclesiastical
causes.
Speaking generally, we may therefore accept it as a fact that
the period of growth in Christian art came to an end about
A.D. 400. It was then that the artless naturalism of this first
period yielded to a new taste leading less to creative work than
to an eclectic treatment of forms which in the first Christian
centuries had grown up for the most part from national seed.
We may note the same thing in the case of other creeds and
civilizations. Indian Buddhism in substance perfected its art
contemporaneously with Christianity, and began its expansion
only a little earlier. In China the period of growth came to an
end about a.d. 800, in Mohammedan countries about a.d. iooo;
in both cases all subsequent art lived upon that of the earlier
period.
II. Church Art oj the Fifth Century
Thus, in my judgement, Christian art starts with the per-
meation of national groups by a spontaneous religious feeling
free from ecclesiastical control ; herein lies the fundamental
difference between my attitude and that of other scholars toward
the views which have hitherto prevailed. It is, indeed, true that
the wall-paintings of the catacombs and the reliefs of the sarco-
phagi have actually been judged in the main by the standards of
a time as yet uncontrolled by the Church ; but this has not been
the case either with Church buildings or their decoration. Yet
every student has now begun to see that before the Peace of the
Church in a.d. 313 there were already in existence both com-
munities, and buildings adapted to their requirements. In spite
of this no one drew the following unavoidable conclusion. The
first three centuries had been so prolific in the creation of types
that the Churches into which at the close of this period the
communities had developed, found themselves confronted with
fixed usages which they adopted without question. These usages,
however, soon acquired a permanent authority through an
original association with the Founders of the Churches, with
individual apostles, with missionaries, or with princes such as
Abgar, Trdat (Tiridates), or Constantine. The Church itself
found it hard to make headway against them when the positive
D 2
/
36 COMMUNITY, CHURCH, AND COURT
necessity for change was shown by subsequent experience gained
by contact with the masses and by liturgical development. The
study of Armenia satisfied me as to this conflict between the
inherited traditions of the fourth century and the demands in
the next century of a Church pressing towards universally valid
forms. I conclude that the course of events in Mesopotamia
was probably identical. We see the same struggle repeated at a
much later time when the pioneer architects of the new S. Peter's
at Rome envisaged a centralized structure with radiating arms,
while the Church demanded a building with a long nave. On
this occasion the original influences which the Church rejected
won in her despite a momentous significance for architectural
development.
It is characteristic of a Church claiming wide allegiance
that it seeks to invest with a validity beyond the natural limits
of place, time, and environment, the architecture and the repre-
sentational art which have approved themselves and become
customary in the chief seat of its activity ; it attempts to impose
them everywhere. This course of action made on the one hand
for consistency in development, on the other it was a source of
conflict. Let me cite a typical example from Armenia. In that
country, during the fourth century, the people had adopted as
its church -type a centralized domed building in complete
independence of outside influence. But when the representatives
of the Armenian church sought union with the now powerful
Greek and Syrian community, and were received with only
too open arms, a type quite different from their own, the long
church without any dome, was declared the one admissible form,
and in the fifth century was actually forced upon them by this
foreign influence. Naturally enough, friction ensued ; only
so can we explain such enactments as Canon 182 in Armenian
Church law (Canon 22 of the Successors of the Apostles), which
runs : * Only the bishop orthodox in faith may design the plan
* of a church, or the Chorepiscopos or the Peredut with the
* bishop's consent. If any presume to plan a church without
' the bishop or Chorepiscopos, we ordain the destruction of the
* plans. Should, however, an unauthorized plan be sanctioned,
' we recommend that it be again submitted for approval. Thus
* shall the designing of the church be blameless.'
The Armenian canons are incoherently arranged, and it
is hard to assign to any particular one either a date or place of
COMMUNITY, CHURCH, AND COURT 37
origin. Canon 182 must in any case be earlier than a.d. 719, the
date of the fifth Council at Dwin. But since it is an apostolic
canon, its substance should go back to the earlier centuries of
Christianity, though it is perhaps unlikely to date from before
the fifth. It is another question whether the canon was from the
first actually followed in Armenian church-building. Agath-
angelos certainly describes S. Gregory laying the foundation
walls of the martyria at Vagharshapat with rule and plummet in
his hands. This passage may afford the simplest explanation
of the nickname given to Nerses III — Shinogh, ' the builder ',
even apart from his active encouragement of construction. Be
this as it may, such points are worth notice as proving the strong
interest taken in the work by the patron, here the Church and
its rulers. But the long congregational church introduced by
the bishops in the fifth and sixth centuries was not constructed
in Armenia in the usual Western manner with a timbered roof,
nor was the barrel-vaulted roof employed after the Southern
fashion. Barrel vaults were used, indeed, but with a central
dome. The efforts of the Mediterranean Church after sole
authority appear as early as the fifth century in the insistence
upon a long nave ; but the consequent disputes led to compro-
mises, which fell out differently in Armenia and in Mesopotamia,
in Syria, and in Asia Minor.
With the official intervention of the Church in problems
hitherto left to individual communities and their architects,
a prescribed taste came into existence ; compulsion was exerted
to give permanence to a solution already accepted : this throttled
creative effort and brought ' styles ' into existence. The above-
cited Armenian canon clearly illustrates this point. In Rome,
research has not revealed facts equally important to the
genesis of Christian art, because in Rome no discord could arise
between an indigenous architectural type and that of an intrusive
foreign Church. Discord was equally impossible in Con-
stantmople and other ecclesiastical centres in Asia Minor,
Antioch, Alexandria, and Carthage. Here the Churches main-
tained active intercourse among themselves, and decorated their
severely practical and simple basilicas with forms which, in so far
as they were widely distributed, developed on the common lines
of Hellenistic art. Nevertheless it is far from being proved
that the timber-roofed basilica, preserved in so many Roman
examples of early Christian date, was their one recognized form.
38 COMMUNITY, CHURCH, AND COURT
The school which followed Ainaloff in ascribing to Alexandria
a general leadership in creation had already deduced the existence,
in the field of architecture, of individual forms more or less
closely related to vaulted building. Moreover, it is b^ no means
certain that the abrupt decline of vaulted architecture in imperial
Rome at the beginning of the Christian building period occurred
to an equal extent in the great cities of the East. The octagon of
Constantine at Antioch, and the same emperor's buildings at
Jerusalem and Constantinople, suggest, at the very least, notable
exceptions, to be explained partly by the persistence of the
Hellenistic tradition as to vaulting, partly by the coming of the
domed and square-planned church of Armenian origin. At
Ravenna, Bishop Ursus seems to have vaulted his basilica about
A.D. 400 ; at Milan, Ambrose probably followed the same course.
An organized Church will always be in favour of placing
restrictions on the progressive artist ; its goal is not freedom but
uniformity. But even in the fifth century and in the Mediter-
ranean area there were several Churches, and their quarrels over
religious hegemony sufficiently explain the general situation.
Further, Alexandria and Antioch had not risen to greatness
because the West and Rome supported them, but because they
had attracted to themselves the trade of the East which brought
with it oriental motives in art. These motives they handed on
to the West, chiefly to Rome and Constantinople, which in this
way received many new ideas at second hand.
The churches of Carthage with their three aisles and double
columns have a tentative character which hardly attests a strong
indigenous art. It is practically impossible to say how far the
initiative may here be due to Alexandria, for ancient Alexandria
is almost completely destroyed, and documentary sources are
few. Judging from the remains of churches above ground and
those laid bare by excavation, we might infer a marked contrast
between monastic and city churches. The former expressed
national aspirations and were in close touch with the art of
North Mesopotamia, Armenia, and Iran. In the cities, on the
other hand, churches were built in or near the ancient temples,
like Thebes or Denderah, and here the timber-roofed basilica
remained the model. The sanctuary at the grave of S. Menas
in Lower Egypt belonged, indeed, so definitely to this Hellenistic
group that Islam was able to provide itself from this source with
columns for its mosques on a liberal scale. But, generally
COMMUNITY, CHURCH, AND COURT 39
speaking, the national art of Egypt won the upper hand with the
spread of Christianity ; it stood closer to the art of Syria- Mesopo-
tamia and Armenia than to that of Greece, and triumphed in the
monasteries against the cities. This is shown alike in the
architectural types of Upper Egypt, and in a very common form
of church decoration imitating in stone the effect of oriental
wall-linings ; the builders were not content to use the ordinary
Greek ornament.
Syria was divided into two parts. The littoral and Jerusalem
attracted all the artistic talent capable of executing the imperial
plans. A widely distributed national art is only found in East
Syria, but with such a wealth of monuments dating from the
fourth to the sixth century that we are well justified by the
general tenour of development in assuming a flourishing art of
the same character in regions where no buildings of the time
are actually preserved, or attested by inscriptions. On the
analogy of Armenia we shall regard precisely this period of three
centuries as of capital importance ; there the indigenous develop-
ment lasted out the fourth century, while the fifth and sixth
centuries witnessed a general attempt to impose the Mediterranean
type of basilica with timbered roof. We are here only concerned
with the native style of architecture in Syria, which apparently
leaned upon Arab tradition and found a permanent home in the
stone country of the Hauran. Here the construction of a church
shows a line of wide rib-arches, all supported on piers projecting
from the wall. Apparently the side walls were carried up as high
as the crowns of the arches and the roofing was then done with
stone slabs. Similar methods prevailed in domestic building.
I need not go into further detail, since from the seventh century
onwards this architecture falls out of the development of Christian
art. But probably the Syrian example strengthened the tendency
to support arcades upon columns, observed quite early in isolated
instances like Spalato and Pompeii, and destined ever afterwards
to remain characteristic of Christian architecture.
Like Syria, Asia Minor also fell into two divisions : a national
central region in the interior, the aflSnities of which were with
Mesopotamia and Iran ; and a littoral, entirely in Greek hands,
and one of the regions from which in Early Christian times
Hellenism derived its principal support. In Asia Minor, more
than anywhere, the Hellenistic spirit came into immediate contact
with another national genius, strongest in those provinces of
40 COMMUNITY, CHURCH, AND COURT
Pontus, Cappadocia, and Cilicia which form connecting links
with Armenia and Mesopotamia. In the coast-lands timber-
roofing prevailed in architecture and representation in pictorial art ;
in the interior, vaulted construction and non-representational orna-
ment. Constantinople reflected both aspects of this two-sided
Anatolia. Down to the fifth century it went with the Hellenistic
littoral ; it then yielded wholly to the East, till iconoclasm
threatened all that remained to Hellenism, now that the timbered
roof was replaced by the barrel- vault and dome. Nothing but
a regular ' Renaissance ' availed to keep Byzantine art apart from
that of Iranian Armenia, a revival which brought back Hellenistic-
Semitic figure art in the decoration of the now predominant
domed church. The spirit of Edessa and Nisibis triumphed ;
orthodoxy adopted painting as its most effective instrument for
influencing the masses ; in this ' Renaissance ' instruction was
everything, and art had little voice. Thus Asia Minor lost with
its dualism its importance for the development of art. Armenians
and Seljuks possessed themselves of the old Ionian soil ; the old
languages, spoken and written, partly died out ; to-day in Anatolia
there is a large Turkish-speaking population of Greek descent.
In the same way the whole of Egypt succumbed to Arabic.
But in the fourth and preceding centuries Asia Minor was
richer than any other land in national characteristics ; achieving
its best in all directions. In the Hellenistic area sculpture was
at so high a level that Praxitelean types were adopted to embody
Christian ideas. It was the same with painting. The Psalter
was illustrated with miniatures conceived in the Pompeian
manner, and marked by an attractive symbolism. Yet in immedi-
ate neighbourhood to this Greek world was the utterly different
world of the Semite and the Eastern Aryan. The custom of
wearing clothes figured with biblical subjects, attested by the
rebukes of Asterius of Amasea, suggests a Semitic rather than
a Greek taste. In the interior of Asia Minor the Mesopotamian
barrel-vault and the dome of Armenia and Iran are the decisive
forms rather than the Hellenistic timbered roof.
We must, therefore, draw a distinction between the inland
regions in Egypt, Syria, or Asia Minor, representing the wide-
spread indigenous element, and the coast-strips with their
ecclesiastical capitals. These are either old metropolitan cities
like Alexandria, Antioch, and Ephesus with its district, or else
temporary residences of Roman emperors, like Ravenna and
COMMUNITY, CHURCH, AND COURT 41
Milan. They are all borrowers ; none are creators, though indeed
their populousness admitted of buildings on a scale unheard
of in the lands actually providing the types, and effects sometimes
attained were such that the original importation is almost undis-
cernible. There was a time when we believed these cities the
creative centres of Christian art; for the period before a. d. 313
we thought principally of Alexandria, for the next period, of
Antioch ; while for later times we gave the first place to Constanti-
nople and to an Asia Minor regarded as purely Greek. The advance
of knowledge with regard to the East has wrought a fundamental
change in these beliefs. The old Hellenistic capitals were in
Christian times emporia for the interior in a double sense. They
did not merely sell their merchandise to countries now re-
awakening to a sense of national life ; they also adopted and incor-
porated Mediterranean art, the architectural forms brought to
their knowledge, a point of much greater importance to art-
history. In architecture the way was thus prepared for the
coming of the domical building, with the dazzling ornament of
internal surfaces. Representational sculpture was deprived of
its former dominant position by the influence of the inland
countries. A new intellectual centre, Edessa-Nisibis, took from
painting its symbolism and substituted a historical and didactic
manner. Until Christianity became the State religion, the
various ecclesiastical capitals retained their leading place.
Let us sum up the argument. Down to about a.d. 400 church
art was represented in the East by independent national groups ;
in the West, there was in architecture a general approximation to
the Hellenistic hall of assembly, which by retention of the
wooden roof, the longitudinal axis, and the gabled fafade, still
more or less preserved the tradition of the ancient temple.
After A.D. 400, a movement began in the East which ceased to
have any immediate connexion with religion. It sprang rather
from the practical brain of Fathers of the Church, just as in later
times the secular art of Constantinople sprang from the predilec-
tions of a Court. The people were at once to be instructed in
the faith and dazzled by magnificent display ; this was the
astute policy of the Aramaean and Greek scholars in Antioch,
Edessa, and Nisibis, and, at a later time, of the Byzantine Court.
Out of it arose the millennial reign of a spiritual and temporal
autocracy in the old Semitic sense, authority in Church and State
claiming to act together in the name of God.
42 COMMUNITY, CHURCH, AND COURT
ni. Autocratic Policy of the Court
Architecturally, the effort on the part of the Church during
the fifth century had issued in a uniform type, that of the wooden-
roofed basiHca. But as early as the fourth century Constantine
had of his own initiative produced a Church architecture unique
in the Mediterranean littoral, and intelligible rather as a continua-
tion of the earlier Hellenistic-Roman vaulted style with important
additions from Armenia and Mesopotamia than as a develop-
ment of the modest and severely practical buildings erected by
the communities of the time. Constantine and his successors
rescued something of the great Hellenistic and Roman vaulted
style for the benefit of the growing Christian movement. Com-
pared with the building done by Churches and communities in
the recognized fashion of the hour, the imperial creations give
many proofs of individuality. Now, as in earlier times, they
challenged the artistic talent of the world ; they displayed in
isolated and conspicuous examples the new forms first carried
westward by the Goths after the entrance of the northern peoples
into the sphere of Christian art ; they developed the vaulted archi-
tecture which we commonly describe as Romanesque and Renais-
sance over an area embracing at least the whole of Western Europe.
But in the East something different grew out of the style adopted
by imperial Byzantine art and the allied art of the Byzantine
Church ; this was the orthodox church type which everywhere
remained unalterably attached to the dome. The two following
chapters will be devoted to a discussion of this rich development ;
our immediate business is to consider the significance of Court
influence on art, especially that of the Byzantine Court,
Criticism makes the first period of great Christian art begin
from A. D. 313, when Constantine abandoned the former Greek State
religion for Christianity. In doing so it fails to see how important
to the development of Christian art was the fact that in this
political move Edessa had anticipated Rome at the beginning
of the third century, and Armenia at the end of it. The older
* Founder's religions ', Buddhism at their head, had followed
the same path : none of them attained a ^reat creative art until
a king associated the new faith with his political ambition, leaving
thereby an inestimable heritage to posterity. But none of them
invented new types of buildings any more than Christianity.
Novelty is not beloved of State religions. Their way is rather
COMMUNITY, CHURCH, AND COURT 43
to enslave the artist in the service of their ambition, and to make
the achievement of his own ideals difficult ; to demand of him
first and above all skilled and willing craftsmanship, and, like the
Churches, to subordinate in him that expression of personality
which is the foundation of all true creative art. The result is
that the indigenous artist dies out, and no record of his activity
remains unless he happens to have been directly under court
patronage. It is the aim of State Churches as far as possible to
keep art uniform in the interest of their arbitrary aims, to cripple
the natural development which comes of competition, and all
the variety created by difference of place, period, and social
environment. The rise of three distinct styles in ancient Hellas^
was only possible in the city States before the time of Alexander ;
in like manner we shall only understand the great development
of Western art in the so-called Middle Ages, when almost every
district in France evolved its own architectural forms, if we
realize the leading part played by individual monasteries and the
advanced civilization of the towns.
The transition from individual believers to communities,
small and large, is relatively easy. A Church first grows intolerant
when she has the State at her back, when she is used as an
instrument by the temporal power. Establishment at first
strengthens a Church, but by degrees it leads to the subjection
of religion by the secular power ; God is constrained to necessities
of State. This sequence of events is easy to follow in the case of
Constantinople. To begin with, the Church alone gained by
the proclamation of Christianity as the State religion ; she alone
took up the battle with local particularism. Justinian was the
first emperor to force both Church and Art into his service.
Though Alexander the Great laid the foundations of absolute
monarchy in the modern sense of the word by his action in
Persia and by his visit to the Temple of Jupiter Ammon, it was
Christian emperors in Constantinople who first succeeded in
enslaving the Church; it was they who began the policy, still
surviving in certain countries, by which she is involved in state
affairs, and thereby divorced from her true spiritual function.
The Church which has accepted the supremacy of a Court
abandons the art which it cannot keep for itself or the art which
it has trained to uniformity ; it must perforce place it at the service
of the ruler who demands an effective parade of power and a dis-
play of dazzling magnificence. When this happens, we begin
44 COMMUNITY, CHURCH, AND COURT
to find interiors gorgeous with costly materials and mural decora-
tion which completely subjugate the devout spectator. Such
interiors spread as an imperial fashion through all parts of
Byzantine territory ; their provision acquired great economic
importance for the capital on the Sea of Marmara. The trade
in marble columns fashioned at the quarries on the island of Pro-
connesos carried such sculpture from Constantinople all round the
Mediterranean and Black Sea basins. It familiarized every region
accessible by harbour, river, and connecting land-route with
the modes of ornament created for imperial use ; it brought
back the riches of all these countries to the capital. If we remem-
ber the imperial monopoly in purple, in silk fabrics, and similar
things, we readily understand why this splendour turned every-
where to the imperial advantage. The decoration and furnishing
of churches became one of the most valuable commercial assets.
Basilicas like that of Parenzo, or the churches which served as
quarries for the building of mosques at Kairwan near Carthage,
or in the Temple Area at Jerusalem, might, all but the bare walls,
have been imported ready-made from Constantinople.
In this way there came into existence a kind of Imperial art,
the Byzantine ; an art economically more productive, though
scarcely more creative, than that of the Roman empire. At
a later period Baghdad in its turn traded the products of its art
from the Indian frontier to the Atlantic ; the beautiful wood-
carving of the mimbar and the glazed tiles of the mihrah in the
above-mentioned great mosque at Kairwan come from Baghdad.
But it would be an error to suppose that Rome, Constantinople,
and Baghdad were anything more than marts for the exchange
of works of art. The pierced carving of the mimbar and the
mihrab tiles of Kairwan are shown by their style and quality
to be of Iranian origin ; the rich decoration of the Proconnesian
capitals in the Ravenna churches or that carved on Egyptian lime-
stone along the Nile valley attests the same descent. Courts
are like sponges ; they absorb everything within reach and
force it into the arteries which nourish their strength.
In that great emporium, Constantinople, two churches rose
as unique examples of imperial craze for splendour ; both of
them go back to Justinian, as did the Church of the Virgin in the
Temple Area at Jerusalem, where Solomon in an earlier day
sought to eclipse the buildings of all neighbouring princes.
Justinian in his turn was not content to excel Solomon alone ;
COMMUNITY, CHURCH, AND COURT 45
he had to outdo his predecessor Constantine, who had shown
how a Christian prince could also erect splendid monuments
in glorification of royal power. The Churches of S. Sophia
and of the Holy Apostles are in a sense only successors of the
structures erected to proclaim the triumph of Christianity. I
will now consider the sumptuous buildings of these two emperors
from the point of view represented in this volume.
Constantine demanded that the basilica of the Holy Sepulchre
should not only be more magnificent than anything elsewhere,
but also so exceptional as to surpass the best that any other city
could show. The chief architect was a certain Zenobios ; if it
is true that he placed a dome over the crossing which served as
a model for all western buildings, then there must have been
two domed structures close together on this site, one over the
sepulchre, the other over the long basilica. Further, a trefoil-
ended building was begun over the cave of the Nativity at
Bethlehem ; a dome was erected over the spot on the Mount of
Olives from which Our Lord ascended, and so on. A whole
series of buildings rose which must have produced an impression
of great individuality, compared with those which we may
conceive to have stood in contemporary Rome ; examples,
conspicuous far and wide, of the resources which the imperial
power of the day could draw indifferently from East and West.
Jerusalem and Syria did not even supply all the materials, for the
columns were assembled from the most various regions. The
effect of the buildings erected by Constantine in his new capital
must have been much the same. The octagon at Antioch,
which was the model of S. Vitale at Ravenna, must have more
nearly recalled the eight-foiled type of Armenia than the original
plan of the Baths over which it was erected. Wherever in the
Mediterranean area imperial policy touched architecture, we find
some exceptional feature in the form of the building. It was
ecclesiastical influence which first substituted the uniformity
of the long plan in the two centuries succeeding Constantine.
Then came Justinian. His architects were from Asia Minor,
but we should seek in vain in their native cities, Tralles and
Miletus, the types which they adopted. The younger Isidore is
heard of at a later period in the district between the Orontes
and the Euphrates ; it looks as if not only he, but also the older
Isidore and Anthemius had in their time studied their profession
on the borders of Armenia and Persia. For the Church of the
46 COMMUNITY, CHURCH, AND COURT
Holy Apostles was just such a many-domed type as is represented
by the palace of Firuz Abad in Persia, and the church once
occupying the site of the Halabiyeh mosque in Aleppo, where
five domes were so arranged as to form a cross. In the wooden
churches of the Ukraine we find the dome similarly used in groups
of three or five even to this day, a point to which I will later recur.
IV. The Church of S. Sophia
This supreme monument of Eastern Christianity has never,
like S. Peter's at Rome, inspired a host of imitations. It stood
in a majestic loneliness until the Turks, in their characteristic
way, once more took up the tradition which it embodied. In
conception the church is purely Armenian ; a central dome over
a square plan is supported by semi-domes abutting on the sides
of the square. It is true that the Armenian niche-buttresses are
replaced on the north and south by vaults between massive
piers which allowed the introduction of galleries ; but the
architects of the Ottoman conquerors were quick to recognize
the original design, and, in the huge mosques built in imitation
of the church, to restore the Armenian four-lobed plan.
S. Sophia, as we see it, results from the demand of the
Church for a long nave even in a building with only a single
dome ; by the introduction of galleries into an Armenian plan,
it meets the need felt alike by Church and Court for a hall of
assembly on an imperial scale. The Armenian plan is impressive,
and the combination of Iranian decoration and Greek organic
structure notably contributes to the almost overpowering effect
of the interior.
Justinian's other churches, both those which survive and
those known by the descriptions of Procopius, all show a like
exuberance of cosmopolitan features. We have already noticed
the Church of the Holy Apostles. S. Sergius and S. Bacchus
at Constantinople is an octagon inscribed in a square, itself
transformed into a square by great niches in the Roman manner,
and surmounted by a * melon dome ' such as we find in Armenia.
S. Vitale at Ravenna has an eight-lobed plan, probably copied
from that of Constantine's octagon at Antioch, on whicn I shall
have more to say. The common run of churches built before
and after Justinian bear an aspect quite distinct from that of
COMMUNITY, CHURCH, AND COURT 47
these buildings. To these churches I will next proceed after
a few words on decoration.
Splendour in decoration was as little known to the old Greek
religious architecture before Alexander as it was to the Armenians
in their earliest churches, or, in much later times, to the architects
of early Gothic churches in the north . The Hellenic spirit expressed
itself by noble sculpture, the mediaeval through effects of mass
and space. The Armenians knew only facings of dressed stone,
with occasional painted ornament in the interior. It was the same
in Syria and Asia Minor. In Mesopotamia and in Egypt, on the
other hand, decoration was richer. These lands of the great
eastern monarchies were the real homes of that desire to enhance
artistic effect by decorative splendour which only the resources
of despotism can satisfy ; the monarch sets the mode, and all^
the world tries to follow it. So the Phoenicians gave Solomon
the idea for his temple, and the temple in its turn was eclipsed
by Justinian's church. If we read the description of the building
written by its panegyrist Paul the Silentiary after its dedication ;
if we bear his account in mind when we contemplate the effect
of the interior, still marvellous to-day despite the whitewash of
the Turks, we shall have to confess that for sheer splendour the
utmost which the Jesuits have achieved falls short of the Byzantine
miracle. AH is resplendent with the most precious material,
from the floor of the church, through the walls and vaults, up to
the crown of the dome. If imagination adds the woven fabrics,
the hangings and the like, the objects in gold and silver, the other
metal work, and the hanging lamps, which once completed the
picture, we may form some conception of the magnificence
attained by this Christian art. Justinian did not, of course,
originate this trend towards splendour ; his part was systemati-
cally to assure its triumph, and to force Europe ever afterwards
to walk in the way which he prepared.
This determination to enhance by decoration an achieve-
ment already good in itself also characterised the representational
art of the time ; in this field, no less than in that or architecture,
it worked in opposition to unadorned truth and to the grandeur
of simplicity. The Court insisted that the subjects on which
the eyes of the congregation were focussed should be imperially
conceived. We remember the two scenes in S. Vitale at Ravenna,
and how in that church the type of Christ had to be made imposing
enough to deserve the homage of a Justinian and a Theodora ;
48 COMMUNITY, CHURCH, AND COURT
we recall the figure throned on the sphere of the world, with
angel body-guards, dispensing grace and receiving worship in
a pose of formal and rigid majesty, his hand resting on the
mystic scroll sealed with the seven seals. We remember the
overwhelming splendours of the remaining mosaics, which seem
to clothe a dogma itself in all the hues of a peacock's plumage.
We have to admit the success of the method. Justinian drew
upon the resources of the whole world that his church-interiors
might fascinate the believer ; his artists wrought the material
with the ease of supreme accomplishment in the service of his
overweening will to power.
We like to imagine that the political centre of Christianity
after a.d. 313 was also the source of the creative spirit in formative
art. But art implies personality. Now a seat of Government
may produce personality in the sphere of politics, of which the
aim IS material domination. To personality of this kind art
seems the best means of decking ambition in the garb of beauty,
just as religion is the best means of cloaking it with virtue and
nobility. But the personality which inspires art is not like this.
It flourishes less at courts than anywhere else in the world. For
at the seat of power everything is subordinated to politics ; the
forces willing to accept this fact are always welcome ; those
which are not willing must either emigrate or remain aloof.
The legends about the erection of famous buildings are instructive
in this connexion. They describe the patron as keeping most
jealous watch over his architect, and even compassing his death
to prevent any repetition of his design.
In centres of political ambition art passes from the hands
of creators into those of actors ; it ceases to be an end to itself,
redeeming us from reality ; it becomes play-acting, a making-up
to please inordinate wealth and power. The surest signs of a court
art are magnificence, and a change in the artists' motive from
expression of feeling to the production of dazzling effect. This
effect must needs rely on monotonously repeated forms and
subjects, in order to mould a public opinion docile to its purpose,
an opinion prepared to remove the control from the individual
creator and to hand it over to the bureaucrat. In this connexion
let us consider the dictum * that Christianity came to be the
Indo-Germanic religion, in which the peoples embodied their
deeper feelings '. In Constantinople Christianity was not such
a religion ; it was more nearly so in Rome, which had been
COMMUNITY, CHURCH, AND COURT 49
freed from the presence of a court. In Rome we shall find among
the surviving mosaics traces which point away from Constanti-
nople back to the centre of East Aryan feeling. In the North,
Christianity remained a blessing to mankind only so long as the
Church did not dominate the State and the State did not find
a wilHng instrument in the Church.
The history of art has made its own path easy by concen-
trating upon finished ' styles ', and by leaving undiscussed all
those early stages of germination which marked the successive
forces at work in nation, Church, and State during the Early
Christian period. It has disregarded the local, historical, and
social movements which forced a material constraint upon the
architect and his work, movements against which the individuality
of peoples or persons had hardly a chance of success. The
measure of development is given on the one hand by the free
forms established by the several communities down to about
A. D. 400, on the other by the trammels and exacting demands of
Church and State. Research will in future have to become more
and more familiar with these distinctions. For me, the effort to
attain such knowledge is the real scientific work, whether it is
possible or not for any single scholar ever to attain it. But
without this aim things pass from a pettiness into a stagnation
which threatens to destroy the importance of our studies. Our
province becomes the hunting-ground of men who either
fail to understand the crucial monuments, or else employ them
to illustrate ideas good enough in themselves, but properly
belonging to other nelds of research. Such was Lamprecht the
universal historian ; such are the archaeologists in their several
spheres. It is high time for students of art to widen their horizon
and get the reins once more into their own hands. They must
not neglect their proper objective to lose themselves in critical
analysis of monuments or texts, in which each will have to take
his part. Problems of form are only a fraction of this.^
* In 1919, while these sheets were in Catholic standpoint, that I may perhaps
the press, I received E. Goller's rectorial hope to see the good results observable
address at Freiburg on the division into at Upsala repeated in circles nearer
periods of ecclesiastical history {Die home. The difference of view is still
Periodisierung der Kirchengeschichte und wide, especially now that I have at-
die epochale Stellung des Mittelalters tempted to give Mazdaism the influential
zwischendemchristlichenAltertumundder part in artistic development which is
Neuzeit). It shows at bottom so clear its due. But the ban is none the less
a realization of my conclusions from the removed, and I can only wish that all
345' E
50 COMMUNITY, CHURCH, AND COURT
I must now indicate the fundamental features in Church
architecture already discovered by Christianity before about
A.D. 400, and ser\'ing as prototypes for subsequent development.
It will be a later task to point out the trammels and exacting
demands of Church and State, in despite of which, after the lapse
of about a thousand years, they finally attained their triumph
in the West. To this task I can do no more than allude in the
next chapter, though it may yet be granted me to devote a part
of my life's work to this question.
who desire a closer familiarity with the to my work on Armenia ; the procedure
problems discussed in these pages should is in marked contrast to that of a hasty
read Goller's address by way of supple- critic more closely connected than GoUer
ment to what I have said here. The with my line of research, whose recent
question ' The East or Rome ' is there pronouncements have only served to
placed in its true perspective in regard confuse the issue.
Ill
Local Variation in the Early Christian vaulted
Architecture of the East
/ARCHITECTURE begins from natural conditions of climate
/-\ and soil ; later it is directed to definite aims, wanders
far afield, following natural lines of communication
between peoples, and satisfying the demands of power and
wealth. In the crucial early period of Christian art there was no
dictation of particular forms, such, for instance, as the old views
on origins assumed in the case of the timber-roofed basilica,
once regarded as the sole point of departure in church building.
Only some world-wide power, such as a State or Church, could
have dictated after this fashion ; Christ, an individual, without
worldly power or influence, assuredly could not have done so ;
and how, in any case, could uniformity have been enforced
during the period of geographical subdivision which followed
His coming } Half a millennium had to go by before Christianity
wielded any such strength either in Rome or in Constantinople,
and even then neither city could ignore all that had come into
existence in the meanwhile. It was precisely the modest first
attempts in Christian art, in all their local divergence, which
really governed the whole subsequent development. It was,
indeed, within the power of Church and State to adopt or reject,
to impose combinations or to fetter by their laws the ambitions of
individual genius. But in the field of art the familiar forms of
the earliest period remained quite as influential as the personalities
of the Fathers or their pagan predecessors in the field of religion.
Since, however, art is subject to less narrow limitations than
language, the sphere in which it is born and works is far wider
not only than that of language but of any other expression of life.
Christ came at a time of intellectual and spiritual satiety,
no less conspicuous in Judaea than in the Hellenistic cities and
in Rome ; it was a state which threatened to engulf all civiUzation
E 2
52 LOCAL VARIATION IN THE
alike in East and West, involving good and beautiful, ugly and
bad, in one common ruin. Hitherto students of art have assumed
that this ruin was actually accomplished ; I myself once so
believed, in so far as I held that at first initiative belonged solely
to the great Mediterranean cities and to Rome, which together
embodied all the creative power given to Christianity in its cradle.
These views were expressed in my book Orient oder Rom, pub-
lished in 1 901. But the experience gained later in Egypt, Syria,
and Asia Minor put me upon the track of other centres inde-
pendent both of Hellenistic and of Roman influence ; in these
was developed an art which became the nursery of Christian
architecture and of its original decoration. The first of these
centres was Persia, the empire next in power to that of Rome,
and, as we have already noted, more tolerant than Rome, so that
in its territory Christianity was openly professed and not exposed
to persecution.
The present is the appropriate place in which to trace the
effect of another cause important in this connexion. The country
then known as Persia included regions which had developed
diverse architectural types, severally adapted to the various
materials at their disposal and old traditional methods of building,
types which flourished less through the patronage of the Diadochi
and their courts than through their hold upon the affections of
the common people. Unfortunately so little remains of pre-
Sassanian times that proof of their existence can only be reached
by retrospective inference.
It was with the utmost difficulty that students of art-history
could be induced to widen their horizon even so far as to include
the monuments of the Greek as well as of the Latin Church.
We may, therefore, expect a long time to pass before they recog-
nize the national movements which took place beyond the Greek
coastlands of the Mediterranean. Against my harmless * Amida ',
they started a regular campaign which can but end in failure.
There is only one course open to the student who sets out not
from preconceptions but from the monuments themselves.
As early as the third century, Syrians had founded a State
Church of their own, with Edessa as its centre. Though this
Church was soon suppressed by Rome, yet neither the archi-
tectural facts recorded in the Chronicle of Edessa nor the
significance of the theological school at Nisibis can be swept
aside or explained away as results of dependence upon Antioch.
EARLY CHRISTIAN VAULTED ARCHITECTURE 53
This school was an independent organization at a very early
date, more open to eastern than to western influence, though
it is true that it afterwards belonged directly to the Greek Church
and was indirectly connected with the Latin through Cassiodorus,
who in his monastery of Vivarium followed the model of Nisibis.
It was, in fact, a Semitic centre flanked by two Aryan regions,
inspiring the movement of art in the fifth century towards
representation, while the Aryans on either side of it had even
earlier assumed the leadership in building and decoration.
This is explained by the fact that the Semitic centre (not
to be confused with the Jewish home of Christianity) already
possessed a fully developed Church when it first made its influence
felt. The Byzantine Court in the sixth century concentrated
all powers in its own hands. In this it followed the example
of the old Semitic monarchies, of which the tradition had sur-
vived, and now helped the Church to arrest the development of
an indigenous and popular art. Yet in the end victory lay
neither with Rome nor with Byzantium, but precisely with this
creative lower stratum. The vaulted architecture of the East
affords definitive evidence of this. It developed a vigorous
growth through the greater freedom which Eastern Christianity
at first enjoyed.
Organized religion is apt to cause persecutions, first at home,
when one form of belief oppresses another, next abroad, with
the rise of state churches at the beck of the temporal power.
The student of art should closely observe both kinds. We have
to remember that until raised to the position of state religion in
the Roman Empire, Christianity was regarded as completely
harmless in Persia, and allowed to expand at will. Persecution
on the part of Persia only began when first Armenia, then the
Roman Empire in West and East, transformed themselves into
Christian States. This explains how the Christian Church in
Persia flourished as early as the second century, and how it is
that we hear, for example, of church buildings in Adiabene
beyond the Tigris at a time when Roman Christians were still
concealing themselves in catacombs, or holding their services
in the palaces of aristocratic converts. This is surely the reason
why church-building in the East should have begun earlier
than in the Mediterranean area. Hitherto we have misunderstood
the whole development because we have always applied to it
the standards made in Rome.
54 LOCAL VARIATION IN THE
The general notion to-day with regard to Christian archi-
tecture resembles that which I held myself in 1901 with regard
to Christian art as a whole, that it originated in the great cities
of the East- Mediterranean coast, especially in the ' cradle of
Christian art ', Alexandria. But this ^ate of Egypt represents
more perfectly than any great Hellenistic city the sterile cosmo-
politanism characteristic of the Late-Roman world as a whole.
Of its own initiative Alexandria never got beyond the timber-
roofed basilica. As long as the regions really inventive in domed
and vaulted construction lay beyond the range of our knowledge,
it was possible to credit Alexandria with the invention of every-
thing. But to-day the case is altered. In what follows I shall
ignore the timber-roofed basilica as a type, in order not to make
this book too cumbrous. A whole library exists devoted to the
subject ; every ' History of Art ' puts it conspicuously in the
foreground. What I wish to do in the present place is to show
that it can be quietly left out of account, and still no fundamental
proof be lacking of the continuity between the architectural
development of the Early-Christian East and that of the Middle-
Christian West (* Romanesque ' and Gothic), or that of Renais-
sance architecture in Italy. The old profession of faith, that
the basilica was the original building used for public worship in
East and West alike, was rashly reaffirmed at the very moment
when plain proofs to the contrary were beginning to emerge
in Armenia. It may be true of the West ; but even on the
eastern coast of the Mediterranean exceptions appear, and no
feature of the timber-roofed basilica penetrated the interior of
Hither Asia except the tendency to lengthen the nave.
The duration of movements in religious art varies within
wide limits, according as creation is free to individuality, or
fettered by some general law. In the first case, though differences
of soil, climate, race, or people may give rise to local varieties,
all can flourish and continue to exist indefinitely, side by side.
But in the second case time will destroy the enforced uniformity
and bring in changes from the world of variety which will end
by undermining the law. In Early Christian times we observe
a notable contrast bet\veen the conditions affecting East and West.
The East starts with a great number of local types, and after
about A.D. 1000 passes to a few fixed ones. The West, on the
other hand, begins with a single fixed type, and after about
A.D. 1000 adopts in succession the two main features which
EARLY CHRISTIAN VAULTED ARCHITECTURE 55
had long co-existed in Armenia, the vault and the dome. If
we survey the distribution of types in the East during the fourth
century, we find in Iran and Armenia the dome over the square
plan, in Mesopotamia the barrel vault, in the Mediterranean
area the timbered roof. Syria alone, in the Hauran where there
was no wood, replaced the timber roof by one composed of stone
slabs supported upon rib-arches.
Such were the types of wide distribution ; by the sixth
century the Eastern forms penetrated Constantinople to the
complete expulsion of the timber- roofed basilica. In the period
down to A.D. 1000, the eastern dome and vault appear in isolated
instances even in the West. But it was only after a.d. iooo that
vaulted construction spread upon a great scale in the West,
the Eastern style, or Romanesque, developing into two branches,
the Northern, or Gothic, and the Southern, or style of the
Italian Renaissance. Romanesque and Gothic are based upon
the long nave ; the style of the Renaissance has leanings towards
domed building, enriching it, however, in an incongruous
manner with features proper to the Greek temple. In succeeding
pages I shall deal with these main groups, adding by way of
supplement exceptional cases found in border-districts, or
transplanted by migration far from their natural centres.
But something more is required than distinction between
the large group and the isolated instance. In the case of the
groups, we must further distinguish between the place of origin
and the area of expansion. In the East it would appear that
in general building-forms originated in the same regions where
they are found widely distributed. But the rule has its exceptions ;
thus it does not apply to the timber-roofed basilica in Syria,
or to the barrel-vaulted church with three aisles in Armenia
and in the interior of Asia Minor. In the two latter countries
groups were formed by adapting the vault to the three long aisles
favoured by the Church ; but vaulted construction had its home
in Mesopotamia and Iran, the long church in the Mediterranean
area. The regions in the south-eastern corner of the Mediter-
ranean, the Syrian coast, and Egypt are of exceptional interest
in this connexion. Here there is a wealth of church buildings,
and in addition to individual structures a variety of groups
represented side by side. But hardly one is indigenous ; almost
all are immigrants originating in other places.
Jerusalem, however, is the chief instance of such immigration.
56 LOCAL VARIATION IN THE
Constantine rebuilt the city as a place of pilgrimage. Did he
avail himself of local building-forms, or did he derive them
from other artistic provinces ? What strikes us most is his free
employment of the dome. This used to be explained on the
theory that he derived suggestions from types represented at
Gaza (temple of Marnas, or Marneton), in Rome (S. Costanza),
or at Nocera. But is it really true that the source of inspiration
was the round building with dome pierced for light at the top
and supported on an inner circle of columns ? Assuredly not.
We find a number of architectural types, some destined to a wide
influence abroad through the mere fact of their association with
the Holy Places, which caused them to be imitated over and over
again. We shall return to this point in connexion with the
Church of the Holy Sepulchre and the Church of the Nativity
at Bethlehem.
In Egypt the state of Early Christian art before the Moham-
medan conquest of a.d. 640 can be more easily understood than
anywhere else. The forms habitual before this date continued
in use ; a further development, such as that which occurred in
Armenia, was impossible. Research, lasting for years, among the
monuments of the Nile valley forced the conviction upon me
that after the extinction of ancient Egyptian art, there was an
end of creativeness ; almost ever5rthing was imported from
Hither Asia.
The old attitude with regard to Early Christian vaulted
building, with a few isolated exceptions, was simply a denial
of its existence. When my Kleinasien appeared, it was declared
that the vaulted churches of Anatolia were an exception to the
rule and to be explained by local conditions. But outside
Germany the true explanation soon began to find acceptance.
The Berlin school and its adherents refused, however, to give
ground, even after the publication of Amida. Far from it ;
despairing efforts are made to this hour to bolster up a worn-out
doctrine. The orthodoxy of the average theologian and the
average archaeologist is too hard hit by the change. Though
the problematic has long been replaced by the certain, people
still prefer the study of phantoms to that of realities. Intelligent
readers of my books on Mesopotamia and Armenia know better
now than to believe the completely vaulted church typical only
of Asia Minor. Another fallacy which must be abandoned is
that which pronounces the absence of the vault in Syrian churches
EARLY CHRISTIAN VAULTED ARCHITECTURE 57
a proof of earlier date, its presence in Anatolian churches a proof
of lateness. As a matter of fact, the whole Christian movement
culminates in the vaulted construction of Central Hither Asia,
as it culminated in the timber-roofed basilica throughout the
Mediterranean area. Christian architecture is linked not merely
with classical forms, but above all with the vaulted construction
originating in Mesopotamia and Iran. If we regard this vaulted
architecture of Early Christianity and the mediaeval vaulted
building of the West as two bridge-heads, the bridge between
them must cross the whole region of the timber-roofed basilica,
which is more or less valueless for architectural evolution. I
have attempted to build this bridge in Kletnasien, in Amida, and
in my book on Armenia. To persist in regarding the facts there
adduced as still doubtful and still requiring exhaustive examina-
tion by specialists is to treat the work of my whole career as
not worth the pains necessary to its comprehension. The
student of art history may learn from this example the attitude
of philology, archaeology, and historical research towards novel
and disturbing views. They can proceed critically and methodi-
cally so long as their especial line of study is not in question
and no one ventures to oppose the articles of recognized belief.
We hear again and again of the ' boldness and freedom of Hellen-
istic architecture ' m connexion with S. Lorenzo at Milan,
S. Vitale at Ravenna, or S. Sophia at Constantinople. But it
would be far more to the point were it recognized that these
churches really represent the expansion of Iranian art on European
soil, an expansion only rendered possible by the complete
surrender of the leading Hellenistic architects to East -Aryan
and Armenian influences.
I have made this short statement of my case by way of
preface, in order that those unfamiliar with the controversies of
recent years may appreciate the valuable material now offered
them, and may not be seduced from the right path if subsequently
they read that the guide who here addresses them is not worthy
of credence.
I. THE DOME
Only one kind of dome is of wide and crucial importance
in the development of church architecture, the dome over a
square bay. The dome over a round or over an octagonal,
bay appears in comparison as the single instance, or as the
58 LOCAL VARIATION IN THE
transient form produced by temporal, local, or social conditions.
It is untrue to assert that local classification is not feasible in
the case of the dome in the East, To begin with, the circular
and octagonal types only appear in the Mediterranean area,
while the dome over the square is of purely Iranian origin. In
Iran it appears with no abutment but that provided by grouping
secondary chambers round a single dome or several domes in
association.^
'^ It was Armenia which in the fourth century first introduced
for use as a church the scjuare building with single dome and
abutment by axial and diagonal niche-buttresses. The long
church lends itself no less than the centralized building to division
into local groups. In Mesopotamia there are the two main
types of long church, one with longitudinal, the other with
transverse nave ; in the Hauran there is the pseudo-vault ; in
the regions still remaining Hellenistic in Christian times there is
the timbered roof ; and this classification does not touch such
further points as the difference between one-aisled and three-
aisled buildings, between those with clerestories and those
without.
When we come to consider other groups built for definite
practical objects such as Baths, which do not require a single
great room for a uniform purpose, but a division of interior
space for differentiated activities, round and octagonal build-
ings at once take a leading place. The Baptistery, related
in purpose to the Baths, is associated with this group, though
it stands for isolation, or at most co-ordination with chambers
designed for other purposes, such, for instance, as those for
^ While the original edition was in dome changed its colour every day of
the press, Erik Peterson drew attention the week (Blochet, Rivista degli Studi
to a remarkable employment of the orientali, iii. 185). It would seem to
dome by the Sabaeans of Mesopotamia, follow from this that among the Ssabians
According to En-Nedim the Ssabians the domed building was connected with
proceed to a chapel constructed of an astrological symbolism not originating
burned bricks and having a domed with themselves, but coming down to
roof ; there they make offerings to their them from earlier times. This makes
God Hermes (?) or Ares : the text at the adoption of the dome in Christian
this point is corrupt (Chvolsohn, Die architecture all the more remarkable,
Ssabier, ii. 37). Perhaps a statement by and the question arises whether there
Akhibar ez-Zemon may be brought into may not have been points of contact in
connexion with this. He relates that the the intellectual history of the peoples,
first Hermes, a figure in the Ssabian The same question is suggested with
cult, erected a lighthouse with a dome regard to India,
at Hermopolis in Egypt, and that this
EARLY CHRISTIAN VAULTED ARCHITECTURE 59
undressing, and those for the gradual introduction of the wor-
shipper to the assembly or the rite, or those at the sides of a
church. There is now evidence for a square baptistery dated
A.D. 359 in the Church of S. James at Nisibis.
The dome over the square is a form which seems to have
developed from wooden construction after the Aryan immigration
into Iran. That country was poor in wood, and their method
of roofing wooden houses with short beams laid across the corners
had now to be reproduced in sun-dried brick, just as we see it
imitated in stone in India and Kashmir.^ Thus a corbelled dome
rose from four squinches at the corners of the square, leaving
a lozenge-shaped opening at the top. The East-Iranian house,
as described by Curtius Rufus, is thus constructed, and even
to-day such units form the greater part of every Persian village.
We cannot at present say whether this type was actually
transmitted to the Christian Church on North Iranian soil ;
the sun-dried bricks weathered quickly, the buildings did not
last long, and ultimately collapsed into heaps of dust. Nor have
excavations yet been made in the crucial area. Yet we may infer
from the ruined palaces of Southern Persia that the development
must have followed two lines ; at first, as at Firuz Abad, domed
cell was merely added to domed cell ; later, as at Sarvistan,
there is a dominant single dome. How far the Fire Temple
may have shared in the development, we cannot at present
clearly see.*
The dome seems at first to have been set upon the four
walls by the help of squinches. When, however, the quatrefoil
plan was introduced, and the soHd walls were pierced by
arches, the spherical pendentive came_ in. Above, in either
case, rose a drum with windows, for the most part octagonal
externally but round within. The windows, one to each face
of the octagon, were originally large, but later became mere slips
splayed inwards.
* As a supplementary note to my book leaves no doubt that its point of departure
on Armenia, p. 631, I may here draw was the triangular area beyond the
attention to a Korean grave, illustrated Oxus, the apex of which is towards
in the Kokka, no. 276, which shows the India. For the portable wooden house
spread of such reproduction to the of the Aryans see Fergusson's History
East; the grave is usually assigned to of Indian and Eastern Architecture, 2 Ki,
King Heigen (rf. A.D. 589). In the interior p. 403.
we see corbelling executed with granite * Cf. Diez, Churasanische Baudenk-
slabs on which are painted beasts and mdler, p. i6.
continuous scrolls, a circumstance which
6o LOCAL VARIATION IN THE
A. The many-domed type.
This type originated in the Iranian dwelling-house. The
palace of Firuz Abad has three domes along the transverse axis
of the whole complex, between the entrance-rooms in the front
and a court at the back. The disposition of the interior rooms
along the transverse axis will demand our attention when we
come to deal with Christian barrel- vaulted structures. This
disposition persisted even after new liturgical requirements
favoured the type with longitudinal axis, though the later Nes-
torian churches commonly have the row of domes on a long and
not a transverse nave. The old church incorporated in the
Halawiyeh Mosque at Aleppo may be of this type, should it
prove to have had three domes down the longitudinal axis. It
was a plan which became prevalent in Aquitaine, a point to
which I will return in the next chapter ; in the present place
I am concerned solely with the main area of distribution, and only
introduce the West in so far as isolated cases prove how far afield
Asiatic vaulted types travelled even though they failed to establish
themselves on any extensive scale.
The Church of the Apostles at Constantinople represented
a combination of the longitudinal and the transverse plans by
disposing the domes along both axes so as to form a cross. This
church was destroyed by the Turks, but the type is copied in
S. Mark's at Venice. S. Front at Perigueux is of the same type,
the central dome at the crossing belonging to both axes, the longitu-
dinal and the transverse. This arrangement has nothing to do
with a second five-domed type, much favoured by the Orthodox
Church, to which we shall return later.
It is an interesting fact that the long church with three
domes in a line, and the cruciform type in which the two Hues
of domes cross each other at right angles are the predominant
forms of the wooden churches in the Ukraine. It may be that
Aryan timber-architecture here passed directly into Christian
church-building ; the persistence of the corbelled dome favours
the supposition. It would be interesting to know why the wooden
churches of Scandinavia show no analogy to these constructional
forms of the South-East Aryans. In Iran the national custom
of building with sun-dried bricks seems to have been transitional
between the old Aryan method of wooden building and the
Christian adoption of burnt brick for churches with several
domes.
I Mastara, domed square with apse-buttresses, c. a.d. 650 ; exterior from
South-West. See p. 61.
EARLY CHRISTIAN VAULTED ARCHITECTURE 6i
B. The one-domed type.
It seems to have been the tomb which even in pre-Christian
times led to the free-standing square building with single dome.
We may suppose the royal Arsand tombs in High Armenia to
have been of this type, as also those of the two founders of the
Christian State in that countr}', Irdat and Gregory, and those of
their successors. The only surviving example is the Baptistery
at Nisibis, near the Armenian frontier, erected in a.d. 359,
probably on the site of the grave of S. James {d. a.d. 338) ; this is
the most ancient ecclesiastical building in Mesopotamia. But
the tomb of Theodoric at Ravenna compensates for what has been
lost. The type did not develop spatially until it had to be
increased in size in order to contain a Christian congregation.
The enlargement was carried out in two ways : by setting several
single-domed units in a row, and by the expansion of the single
unit. The necessary abutment for such martyria was secured
in Armenia by placing great niche-buttresses at the ends of the
axes ^ to receive the thrust of the dome. There are two principal
forms of the square ^^kn with niche-abutment. One, which
I shall call the ' jiiche-buttressed square V shows the four niche-
buttresses projectm^-ffeffl-thc-fOur walls, whiclL stand free as
a visible square ; the other, to be called the 4uatrefoil} shows the
square plan only in the interior, since the nicne^lSuttresses meet
each other at the corners without any intervening wall-space.
Each type can be subdivided into further distinct varieties :
a. Subdivisions oj the niche-buttressed square. These are of
three kinds : that with niche-buttresses on the axes only ; that
with niche-buttresses on both axes and diagonals ; and that
without any niche-buttresses, the dome resting on interior
supports.
I . Niche-buttressed squares with the buttresses on the axes only.
This, the most important variety, was unknown to us until
quite recent times. The first photographs of the simplest
surviving example were taken during the expedition of the
Institute which I founded in connexion with my professorial
chair at Vienna ; this example is the Church of Mastara, dating
from about a. d. 650 (Fig. i, opposite). Except for the band of
arcading beneath the restored roofs and the arched mouldings
above the windows, it is without elaborate ornament ; it finds
* i. e. in the middle of each of the four sides.
62 LOCAL VARIATION IN THE
a parallel in the Church of S. Gregory in the Haridsha monastery,
which perhaps dates from the sixth century. A second variety
is preserved in the Cathedral of Artik (Figs. 2-3, opposite) ;
in this both exterior and interior are richly ornamented with
blind arcading on engaged shafts ; a later example is the Church
of the Apostles at Kars, erected in the second quarter of the
tenth century. This last has blind arcading only round the
dome ; but at Artik the decoration is such as to arouse
the liveliest interest of the Western traveller. It covers two of
the niche-buttresses ; the shafts are surmounted by cushion
capitals, while triple interlaced bands are repeated on the
cornices. The exteriors of these indigenous Armenian buildings
produce an effect of strength and mass, mainly through their
pyramidal form, the roofs of the four niche-buttresses leading
the eye upwards to the dome crowning the structure.
This variety had a wide distribution beyond the limits of
Armenia. We need feel no surprise to find it represented in
North Mesopotamia (Deir es-Zaferan), as well as in Georgia.
The richly ornamented niche-buttressed square at Amman in
Moab is to be connected rather with the architecture of Iran
proper. Here the niche-buttresses are square at the base and are
carried over into a round plan by squinches, as in many other
not purely Armenian examples. The point most important for
us to notice in regard to relations with the West is that the
niche-buttressed square can be traced across the Dobruja
{Tropaeum) into the Czech country, where some of the most
ancient national monuments are small-domed churches of this
kind.
By the side of these national examples we may notice an
adaptation of the apse-buttressed square due to imperial influences
at Constantinople, perhaps immediately to the Court architect,
Trdat, an Armenian, engaged upon the restoration of S. Sophia in
A.D.989. It is to be seen in splendid monastery churches, where
we mark a tendency to replace the niche-buttresses by barrel-
vaulted members. The best-preserved example is the Church
of S. Luke of Stiris in Phocis between Helicon and Parnassus ;
next to it that of the monastery of Daphni near Athens ; in the
third place, that of Nea Moni on Chios, in many ways the nearest
of all to the original type. The intrinsic splendour of decoration
and allusions in texts lead us to conclude that these buildings
were erected under imperial patronage.
2 Artik, cathedral, domed square with apse-buttresses ; South-West view.
See pp. 62, 95, 117, 146.
^■IbT,
a.'i.-i ^
3 Artik, cathedral ; plan. See p. 62.
EARLY CHRISTIAN VAULTED ARCHITECTURE 63
2. Domed churches on square plan with both axial and diagonal
niche-buttresses (i. e. with niche-buttresses both in the middle of
each wall and at each corner). The niches at the comers are
three-quarter cylinders, perhaps originally intended to strengthen
the abutment of the dome, but soon pierced so as to give access
to subsidiary chambers near the eastern niche required for
liturgical purposes. These conditions are still apparent in Vas-
purakan, where the Church of Achthamar (a.d. 915-21) seems to
represent a later example of the original type. In Georgia there
are at least two good surviving examples, the cruciform Church
of Mzchet (about a.d. 600), and the Sion church at Aleni (about
A.D. 1000), the creation of an Armenian architect, where the
niche-buttresses with the rectangular comer-chambers are dis-
tinguishable even on the exterior. The model of the type is
the Hripsimeh Church, and such complete examples are only
to be found in Armenia : Awan, built about a.d. 570, has round
corner-chambers, the Hripsimeh Church itself quadrangular,
though here these and all the niche-buttresses are invisible
from outside, since the whole structure is concealed within
quadrangular external walls above which the dome towers on
its windowed drum. In this one-domed group there developed
a special Armenian form of external church decoration, affecting
the niche-buttressed square ; this consisted of triangular slits
designed to give salience to the niche-buttresses by vertical
furrows of deep shadow, and comprised within a gable which
gives the mass of the roof an essentially new aspect. In the
interiors of this type the barrel vault is employed in connexion
with the corner-chambers and the external walls ; it is inter-
polated between the square central bay and the niche-buttress,
and serves also to accentuate the east-west direction of the main ^
axis. This type is the most singular which occurs on Armenian ^.^^
soil. It has found no favour beyond the Armenian border,
unless S. Peter's at Rome can be regarded as an example.
3. Dome over square bays with axial niche-buttresses and
central supports. This variety has to-day only one representative
in Armenia, the Cathedral of Bagaran, erected a.d. 624-31. In
the inscription, the date is reckoned by the regnal year of Chosrau
Parviz, and further, the oviform arch found with both the pointed
and the round varieties betrays a Persian influence. Four piers
originally supported a dome now fallen in, but once rising from
squinches. High barrel vaults are interposed between these
:r
64 LOCAL VARIATION IN THE
piers and the niche-buttresses. The interior decoration is confined
to narrow geometrical fillets, that of the exterior to arched
mouldings over the windows, and a toothed motive on a slanting
surface, such as at a later time we so often see used as a border
at Venice.
The Bagaran type is known to have existed in Iran and on
the route from Iran to Armenia. One example is seen in a corner
hall, it must be admitted, of rather late date, in the palace of
Sarvistan in Persis ; another is the now-destroyed praetorium of
the second century at Mismiyeh in Syria, where it is modified in
a Hellenistic sense ; a third is a church of the second half of the
sixth century before the walls of Rusafa near the southern border
of Armenia. The Armenian example at Bagaran is the only
one in a region where the evolution of the type can be logically
demonstrated. We conclude that a suggestion originating in
. Persia found its development in Armenia.
Now in A.D. 806 this type appeared on Prankish soil in the
xZ Church of S. Germigny-des-Pres ; it was probably, in the first
^; instance, transmitted to France by Persian or Armenian archi-
tects connected with the immigrant Goths from the Black Sea
region. All the arches in this church are of the horse-shoe form,
and the stucco decoration has Iranian features. The mosaic in
the apse, with the Ark of the Covenant between angels, seems to
find a parallel at Tekor in Armenia. A second example of the
niche-buttressed square with interior supports is S. Satiro at
Milan, erected a.d. 879, though here columns take the place of
piers, and there has been frequent reconstruction. The type
also exerted no unimportant influence on the drawings of
Leonardo da Vinci, who may have become acquainted with it in
Armenia itself. In general the West can only show a few scattered
instances. In the east of Europe, on the contrary, this Greek-
Cross plan with square central bay is of extraordinary importance,
I / though it must be admitted that the niche-buttresses, as seen at
■^V / Bagaran, are here generally omitted. It became the almost
f \ universal type of the Orthodox Church, and as such will receive
fuller attention below.
b. Buildings composed oj niche-buttresses only. The large
niches originally introduced in Armenia to buttress the four walls of
the square-domed unit here touch each other without intervening
walls, and enclose an (ideal) square, hexagon, or octagon ; thus
arise radiating types with four, six, or eight niches in juxtaposition.
V
EARLY CHRISTIAN VAULTED ARCHITECTURE 65
The dome rests on the wedge-Hke projections of the walls between
every two niches ; it has a drum with windows, and a hemi-
spherical cupola, covered by a pyramidal roof. We found
numbers of such structures dating from the Bagratid period at
Ani and Chtskonk, others, dating from the seventh century, at
Agrak, Irind, and Eghiward. The type is probably one used
for Baths, and must go bacfe^ to pre-Christian times. I have
adopted the name quatrefoil ,ior a building of this kind with
four niches. A quatrefoil may be either simple or have an
ambulator}\ In the West the first variety is to be seen at Biella
and Galliano in North Italy, and at Montmajeur (Church of
S. Croix) in France ; an example upon the road between East
and We§t is to be seen on the Vistula at Cracow. The second
variety has more importance. The most remarkable example ^^ ^/)
in Armenia is the sepulchral chapel of S. Gregory at Zw-arth« - ■^■'^^ !j3
notz. erected, together with the adjoining palace, by the Catho-
likos Nerses III about a.d. 6^0 (Fig. 5 , facing p. 66) ; here
my assumption is that the tomb was enclosed m a containing
structure. Excavation proved the existence of a great quatrefoil
with ambulatory, clearly imitated in the ninth century at Bana
and Ishchan, and in a.d. iooo by King Gagik at Ani. In this
group the quatrefoil is enclosed in a circle or a polygon. Armenia
may, however, have possessed buildings in which one quatrefoil
enclosed another. A connecting link exists on the road towards
the west, in the so-called ' Red Ruin ' at Philippopolis ; a
quatrefoil enclosing a quatrefoil is still preserved in the much-
reconstructed Church of S. Lorenzo at Milan.
This example well illustrates the importance of Armenian
discoveries as throwing light on the early growth of Christian art
in Europe. S. Lorenzo has always been regarded as an enigma ;
it has been too seldom observed that Italy can show a parallel
case in the * Minerva Medica ' at Rome, probably erected in
conjunction with a villa of Licinius Gallienus (a.d. 260-8), and
in the time of Constantine buttressed by two great niches.
S. Lorenzo is a niche-buttressed square ; but the Minerva
Medica is composed of ten niche-buttresses once supporting
a dome upon a drum with windows. My impression is that
this building, so abnormal for Roman imperial times, may be the
work of Armenians, who were strongly represented in Rome. If
so, this would be another of the proofs compelling us to admit the
existence of Armenian domed buildings in variety in the fourth
66 LOCAL VARIATION IN THE
century, and simple niche-buttressed buildings even earlier.
The hypothesis will give rise to much investigation ; at the same
time it should provide an unhoped-for solution of the difficulty,
insoluble as long as all attention was fixed on the Mediterranean
area without a thought for the comparative material in the East,
above all in Iran. Armenia now provides the connecting bridge
- by bringing the niche-buttress into the problem.
V \ < S. Sophia at Constantinople is itself related to the quatrefoil
Vy" group with ambulatory. In their imitations of the building, the
^ architects of the Turkish conquerors left out the galleries ; they
placed their great niches on the diagonal axis, and so produced
a regular quatrefoil. The amazing splendour of the interior
decoration of S. Sophia, in which antique and Iranian elements
blend, makes it difficult to distin^ish the underlying plan ;
but the logical affinities of the buildmg now become clear. The
nave, with its dome and two semi-domes, i.e. niche-buttresses, is
so conspicuously Armenian that there can be no doubt as to
the source from which the Anatolian architects derived their
plan. Their problem was to build a Court church which should
at the same time serve as an imperial hall of assembly ; the
galleries were for the accommodation of women.
The second variety with contiguous niche-buttresses, the
'vV six-foil (Fig. 4, opposite), was widely distributed in Georgia,
' . where it underwent the strangest transformations, one of which,
represented at Kumurdo, has been repeated in a striking manner
in the Attic monastery of Dau, though here too a gallery has been
introduced. The chapel in the citadel of Marienberg at Wiirz-
burg, ' the oldest church in Germany ', is a six-foil of the
Armenian kind ; exact investigation of this chapel will show
whether it may not prove to be a veritable landmark in archi-
tectural history.
The third variety, the eight-foil, is familiar through the
example of S. Vitale at Ravenna. It is true that here, too, we
(^/Xmust bear in mind the universal absence of galleries in Armenia,
^^*^^the country where the type is most widely represented and
connected by the most numerous links with other church forms
logically leading up to it or issuing from it. The variety with
galleries makes its appearance as soon as the Greek area is
entered ; we see this as early as the fourth century in the case
of Constantine's octagon at Antioch, erected on the site of earlier
Baths. This building, described by Eusebius, and often mentioned
4 Ani, Church of S. Gregory, mid-tenth century, sexfoil
plan with triangular slits ; South-East view. See p. 66.
...n^
5 Zwarthnotz, Church of S. Gregory, c. a.d. 650,
quatrefoil plan with ambulatory, and palace, p. 65,
EARLY CHRISTIAN VAULTED ARCHITECTURE 67
by other writers, forms a link between Armenia and S. Vitale which
has hitherto been regarded as, more or less, a copy of the Antioch
octagon ; it would appear to have been an eight-foil with
ambulatory. The simple eight-foil occurs at a late date in
S. Michel-Entraigues near Angouleme. The idea of the type
was taken up once more by Brunelleschi in the ' Tempio degli
angeli ' begun by him at Florence, and by Leonardo in his
drawings.
II. THE BARREL VAULT
We have already had to deal with this vault in discussing
domical construction ; as far as Christian Church architecture
is concerned, it seems to have originated in Mesopotamia rather
than in Iran or Armenia. In Mesopotamia there is a whole
series of churches vaulted exclusively in this style. Since
examples of this method of roofing upon the most impressive
scale are preserved in Sassanian palaces, for example in the
celebrated Tak-i Kisra, it may well be indigenous, and have
attained a wide development with the period of Parthian influence.
At first sun-dried bncks were doubtless used, and the earliest
examples will only be found by excavation ; the examples which
now stand are of burnt brick, like the above-mentioned Sassanian
palaces, or, as in the old Assyrian North, of stone and brick in
combination.
A Parthian building, the palace at Hatra, in the desert
between Tigris and Euphrates, about their middle course, shows
a ground floor with barrel vaults only, constructed of rubble
faced with stone ; the Arab method of roofing, as known to us
from the Hauran churches, is only employed in the upper
storeys. If we look back to yet earlier times, the uniform employ-
ment of the barrel vault in Assyrian and Babylonian palaces
may be disputed, but the vault itself can be shown to have existed.
Recent excavations have laid bare temple sites which may perhaps
be considered prototypes of later church buildings.
A. The church zoith transverse nave. There were two regions
which did not follow the Mediterranean area in beginning with
the long church, but opposed to its advance a building remark-
able for its breadth. One is East Syria, where a type originating
in Arabia prevailed. Here we find a line of rib arches supported
on piers ; the surrounding walls are carried to the height of the
crown of the arch, and the roof is formed of stone slabs ; a series
Fa
68 LOCAL VARIATION IN THE
of such arched units forms a long hall. The Mesopotamian
method was different. There the church consisted of a single
transverse nave at right angles to the line connecting entrance-
door and apse. To the occidental, the appearance of the exterior
is unfamiliar, since on entering you have before you not the
gable, as in the case of a Greek ternple, but one whole slope of
the roof, as in Chinese buildings. The best surviving examples
are the Church of S. Jacob at Salah (Figs. 6-7, opposite),
several monasteries in the Tur ' Abdin, and at Kesum near Edessa.
The type may be connected with the church-plan of the early
third century, when Christianity was for a while the official religion
in Edessa, and is perhaps directly based on ancient oriental
methods. It had no influence on the further development of
Christian art, unless it be on the formation of the transept,
especially in that Armenian variety called by me the domed
transept, where a central dome is flanked by niche-buttresses or
semi-domes ; this variety is admirably represented in the Church
of the Virgin at Khakh in the Tur 'Abdin. S. Sophia at
Constantinople shows in so far a distant affinity to this Meso-
Eotamian type with its Armenian dome and flanking niche-
uttresses, as it also has a domed nave ; though this follows and
does not cross the longitudinal axis of the church. But as
suggested above, the more direct affinities of S. Sophia are
probably with Armenia. In the West the introduction of the
transept may have been suggested by the Mesopotamian church
with transverse nave.
B. The long church. The dome and the barrel- vaulted
transverse nave dominate the oldest Christian church-building
of the East. It is possible that a single-aisled long church
existed at the same time. But in the fifth century the favourite
Mediterranean form, the three-aisled long church, began its
expansion, as we see both in East Syria and Armenia. In
Syria, builders at first attempted to retain the old style of round
arches and roof of stone slabs ; but here the timber-roof gradually
gained in popularity. Barrel vaulting now first occurs in Armenia,
probably through the influence of North Mesopotamia and
Cappadocia. In these countries the three-aisled long church
is almost always vaulted ; the timbered roof forms the exception,
though it is found in parts of the Euphrates valley bordering on
Syria, and examples occur in towns as far as Meiafarqin. But
in the monasteries of the Tur 'Abdin in Armenia, and at Birbir
6 Salah, Mar Yakub, church with transverse nave. See p. 68.
7 Salah, Mar Yakub, plan. See p. 68.
8 Ereruk, basilica with ruined barrel-vaults, and fagade towers.
W4 1 I I 1 t 1 1 I I r
9 Ereruk, basilica ; plan. See p. 69,
EARLY CHRISTIAN VAULTED ARCHITECTURE - 69
Kilisse, only the vaulted hall-church is known, with, here and
there, the regular vaulted basilica with clerestory. All the three
varieties of long church are represented in the East by more
or less numerous surviving examples. I am here considering
only vaulted long churches, and omitting all three-aisled buildings
with timber roofs. The one-aisled and three-aisled long church
without clerestory, the hall-church, is widely diffused. Examples
of the regular basilica with clerestory are rare. There is a single
brilliant example at Ereruk (Figs. 8, 9, 19, facing pp. 69 and 91),
as if to prove that the Armenian architect, whose proper interest
was in the domed church, was capable of excelling in the foreign
form also ; the church is a regular basilica, homogeneous through-
out, with a developed western facade between two towers. The por-
tico or narthex at the west end, the aisles, and the high nave wall
above the windows all have barrel vaults ; and the lateral cham-
bers on either side of the apse, which have more than one storey,
have ascending barrel vaults of a type occurring in Asia Minor
and at Como and Aix-la-Chapelle. But in all likelihood Ereruk
is as early as the fifth century, when the Greek and Aramaean
Churches sought to impose their types upon Armenia. Thus here,
too, the theory is confirmed that the most important achievements
in Early Christian art belong to the period between the fourth
and sixth centuries. Ereruk ranks among the most consummate
examples of the vaulted basilica in existence. The distribution
of the vaulted long church will be discussed in the next chapter,
which will be entirely devoted to this subject.
III. ABUTMENT OF THE DOMED BUILDING
Despite the intrusion of the long vaulted church, Armenia
remained true to the dome. Everywhere within its own area,
and beyond it in Constantinople, Salonika, and Egypt, it imposed
its unit, the dome over the square bay on forms transitional to
the long church, using niche-buttresses on trefoil-ended churches,
whether single- or treble-aisled, omitting niche-buttresses but
adding galleries in the so-called domed basilica.
A. The Trefoil (Trikonchos). Trefoils with single nave
result when the western buttress of a niche-buttressed square
or quatrefoil is replaced by a longitudinal or transverse barrel
vault. The last variety, examples of which are preserved at
Khakh on the upper course of the Tigris, and in monastery
70 LOCAL VARIATION IN THE
churches in Egypt, may descend, as already observed, from the
Mesopotamian transverse-naved church, the barrel vault being
replaced by a central dome with flanking semi-domes. The nave
with trefoil end may therefore have a source other than the
cemetery chapel, the derivation suggested in Orient oder Rom ;
it may have originated independently in Armenia from the
niche-buttressed square when the needs of the Church called for
a long hall. This variety is represented in that country by
examples belonging to different centuries, and it migrated
westwards both by land and sea. The cellae trichorae of Southern
Russia, those in the Balkans, and in Italy, above all in Lombardy
and France, are like milestones marking its path ; on the Rhine
we find the tendency to develop three aisles, already pronounced
in Armenia. That country still possesses two large cathedrals of
this rich variety dating from the seventh century, Thalin (Fig. lo,
opposite) and Dwin ; in both the dome is placed in the
middle, in contrast to the usage obtaining where there is only
a single nave. At Kutais in Georgia, the first country into which
Armenian types spread, and in the Mesopotamian region of the
Euphrates and Tigqs, we find such buildings observing this
essential law of the single-domed structure ; for example, the
long church with four niches and ambulatory at Resafa, and the
Church of the Virgin at Amida ; it is not certain, however,
whether in these cases the domes were of a permanent character.
B. The domed basilica. This type is exclusively confined
to Asia and the Balkans. It first appeared on the confines of
Armenia, moving South, and West as far as Salonika. It owes
its pecuHar character to the fact that the Armenian single-domed
church, on entering the Greek area, had to meet the demand
for a women's gallery. The consequence was that the lateral
barrel- vaulted members were modified, as was the case with
the lateral niche-buttresses of S. Sophia, and two continuous
aisles resulted, at any rate, on the ground level.
At Meiafarqin, once, under the name of Tigranocerta,
capital of Armenia, the seventh-century Church of the Virgin
still recalls the Mesopotamian church with transverse nave ; it
is the type of S. Sophia at Salonika, which is an isolated instance
decorated in the Byzantine manner ; it proves the westerly
migration of Armenians and Mesopotamians in the period
before Justinian. Qasr Ibn-Wardan, in the Euphrates region,
and Khoja Kalessi in Cilicia, show the pronounced long type
a,
■*-»
C
0)
T5
.^ o
■•-' Q-t
3
-^ CO
■4-* ,C^
c '*^
1> 1)
c -a
V u
> ^
05
to
-a
J2
c3
ON
N
d,
<o
CO
■>
W
I
U
O
0)
a.
•a
CJ
S
o
6
o
3
B
O
"ft&iTneei
?•
O
O
Q
<
■a
u
■4-»
o
EARLY CHRISTIAN VAULTED ARCHITECTURE 71
co-existing with the transverse as early as the period between the
fourth and sixth centuries. The type soon spread through
Western Asia Minor, and has perhaps two representatives in the
Balkans (Philippi and Pirdop), as early as the Salonikan example.
In their present ruinous state these long churches suggest two-
domed buildings, as does the Church of S. Irene at Constanti-
nople. They, too, were probably also long domed basilicas,
with galleries, S. Sophia at Constantinople, which I have
already noted as an exceptional building, itself belongs to the
series of domed basilicas ; it is a quatrefoil transformed on basilican
lines. Its galleries are a characteristic feature ; had these been
omitted, as they are in the Turkish mosques, the domed basilica
with continuous aisles on the ground level would have been
impossible. The name ' cruciform domed basilica' might be
introduced for domed buildings which have a definitely basilican
appearance, and have adopted, wherever possible, certain charac-
teristics of the basilica such as a striking length of nave, and
corresponding lateral aisles (cf. Odzun in Armenia and Mokvi
in Georgia). The domed basilica at Meriamlik approaches most
nearly to a Hellenistic style.
C. The cruciform domed Church. All transitional forms
were finally superseded by this purely Armenian type. It
developed either directly from the niche-buttressed square with
interior supports (e. g. Bagaran), through the dropping of the
niche-buttresses and a slight increase of nave length (e. g. Mren),
or by the addition of a dome to the hall-church with piers (e. g.
Tekor), the old long barrel vault and the new short transverse
ones by themselves providing adequate abutment. The essential
feature is that in the domed cruciform church, the dome rests
upon four piers at the corners of a square bay and its thrust
is taken by four barrel-vaulted limbs at right angles to each
other and following the two axes of the cross. The result is
a cruciform room with a dome rising from the crossing ; in
Armenia the dome is always in the middle, in other regions into
which the type spread, either in the middle or nearer to the apse,
according as the idea of a long church or that of a domed structure
prevailed.
The type passed in the first instance to Asia Minor, and then,
with the Armenian Dynasty, into Constantinople. The first
emperor of the Hne, Basil I (a. d. 867-86), built his palace-
church, the so-called Nea, on this plan. The type subsequently
72 LOCAL VARIATION IN THE
became dominant in Byzantine architecture, going wherever
Byzantine influence extended, and it has to this day remained
the consecrated form in the countries of the Orthodox Greek
Church. I will not here follow up its distribution, because my
immediate subject is the origin of Christian church-building
in the West. In general it may be said that the triumph of the
cruciform domed church meant the triumph of domical con-
struction. The architects of the Renaissance were the first to
bring about this change in western building. Deferring this
point to the next chapter, I now proceed with the eastern types.
D. The domed hall-church. When the long-naved church
was imposed upon Armenia in the fifth century (p. 7), the dome
was added, and a type produced which bore in itself the seeds
of Christian northern art in Europe (Gothic). Long ago old
controversies as to the origin of the opus francigenum Ted to the
notion that abutment by niches formed a first step in this direc-
tion, though no one had the slightest idea of the facts now known
with regard to Armenia. When the Armenians wished to
transform the long church, giving it at the same time a dome and
spatial unity, they abandoned three aisles for one, and applied
the piers supporting the dome directly to the outer walls ; the
aisles now survived only as recesses without independent function
and spatially belonging to a single domed hall, as was the case
with the much later Barocco church. The piers appHed to the
interior walls have really the same significance as the Gothic
external buttresses ; both are equally supporting piers.
A fully developed example of the domed hall-church is
preserved in the cathedral of a. d. 668 at Thalish (Arudsh,
Figs. 12-13, facing pp. 72 and 73). A building of this period is
unlikely to be the archetype ; since the type affords the one
satisfactory solution, we should expect the Armenian architects
to have discovered it at the end of the fifth, or during the sixth
century, when the conflict between national and ecclesiastical
influences was at its height. The domed hall-church always
retained the dome over the middle of the building. One of the
most remarkable achievements in this style, the Cathedral of
Ani (Figs. 11 and 21-23, facing pp. 71, 94, and 95), interposes
very narrow arches between the piers bearing the dome and
the wall, so that these dome-supports stand free. It is a delight,
in a church earlier than a. d. iooo, to see the builder, the Court
architect Trdat, carrying Armenian art so logically and so
rj
N
a,
o
J3
C4
3
o
CO
_a"
u
u
13
u
a
o
■a
oo"
o
o
J3
CO
>
u
O
C
J3
EARLY CHRISTIAN VAULTED ARCHITECTURE 73
successfully past ' Romanesque ' to ' Gothic ' that many have
been at a loss to explain this cathedral in any other way than
as a reconstruction by a western master in the thirteenth cen-
tury. The careful study of Armenian art has now proved
that the buildings of Trdat in the last quarter of the tenth
century uniformly show clustered columns and pointed arches ;
they still keep, indeed, the cushion capitals, but give the bases
of the colonnettes such deep mouldings that we again receive
a ' Gothic ' impression. If the above-mentioned Canon 182
of the Armenian Church had not interfered with free competition
among architects ; if the Armenian builder had not been too
closely confined to his walls by his technique of rubble faced
with dressed stone, his northern style might have yet more
definitely pursued the methods which the western masters
adopted or rediscovered down to the time of Vignola.
The Gesii, the domed hall-church of Vignola, the first
church of the Jesuit Order in Rome, finds its proper place at the
close of this section. We may at once compare it to Thalish,
a building nine hundred years older ; the only differences are
that Vignola's dome is nearer to the apse than to the west end,
while the lighting of the upper part of his church is not confined
to the windows under the dome. Vignola set his dome on the
axis of the church near its end, and lighted the nave throughout
its length from above ; while the Armenian always jealously
preserved that unity of space and lighting which belongs to the
very nature of the dome. We shall return to this church in the
next chapter, when we discuss S. Peter's at Rome.
Let us now resume the contents of the present chapter.
It conveys a false idea of development in church-construction
to assume that Christian ideas were first embodied in classical
forms, and then diffused these forms throughout the world.
The precedence habitually given to the Greek and Latin languages
is no less misleading than that accorded to classical architecture.
Aramaean, Iranian, and Armenian deserve, to say the least, an
equal, if not a privileged position, since it was they which provided
the real stimulus to development. If this is granted, we can better
understand how it was that when Diocletian reconstituted the
Roman state, he had no choice but to substitute a Persian for
a Roman organization. The changes introduced by this emperor
did not affect the social structure alone, but equally the whole
fabric of art. If this truth was overlooked, and people refused
74 EARLY CHRISTIAN VAULTED ARCHITECTURE
to admit the force of my contentions, this only showed that in
the study of Early Christian architecture group-distribution and
sporadic occurrence had not been properly distinguished. It
showed, above all, that to this very hour students have allowed
themselves to be prejudiced against my pioneer work by the
academic views so obstinately retained in certain European
capitals. In scholarship there are still unscientific influences at
work convinced of their power to arrest the advance of learning.
It must not, however, be supposed that the question of 1901
* The East or Rome ' still has the old importance to-day, or that
a full agreement might be reached by changing the formula
to ' The East and Rome '. There has never been any question of
splitting hairs in this fashion. As far as architecture is concerned
it should now be superfluous to add a single word. After the
fourth chapter, dedicated to the West, we shall be in a position
to judge whether the hypothetical element in our attitude towards
decorative and purely representational art has not gained support
by the certainty gradually attained in the field of architecture.
For in this field we are no longer dealing with mere suppositions
set up just to test the point whether things might not have been
ordered in a certain way, but with facts which cannot but produce
their eflFect upon a scholar's mind. These facts no longer allow
men to persuade themselves on the authority of mere texts that
things ought to have happened in this or that way. The specialist
who investigates essential facts and their development by the
comparative study of monuments has firmer ground beneath
his feet than the philologist who presumes to compose history
from written sources, and thinks by this means once more to
confine the whole development to Roman ground. As soon as
it becomes a question of artistic creation, Rome has nothing to
say in the matter ; we shall see an example of this in the mosaics
of Roman apses, on which I have something to say in Chapters VI
and VII.
IV
The Succession of Periods in Western
Architecture
IN accordance with the chronology above suggested, I do not
regard Christian architecture, the origin of which fomis my
subject, in the narrow ' Early Christian ' sense, but from
a wider point of view, passing beyond the Middle Ages to the
High Renaissance, even to the beginning of those influences
diffused over the whole of Europe by the Counter Reformation.
The usual division is as follows : Early Christian period, down to
about A. D. 467 or a. d. 568 ; period of the great migrations, down
to Charlemagne ; Ottoman period, down to about A. D. 1000 ;
lastly, the Romanesque, Gothic, and Renaissance periods. But
this classification is only justified if confined to the narrow
sphere of western civilization, and even so, it can only be admitted
as part of the machinery of historical research. It becomes
untenable as soon as the horizon is enlarged in the manner
proposed in these pages. What really dominates all these
centuries, considered from the point of view of art, is the struggle
between East and West, or that between North and South ; in
any proper classification this fact must be recognized, and in
naming its several divisions we must take it into account.
In the previous chapter I purposely disregarded the basilica.
I here in like manner disregard the survival of antique tradition
in such familiar architectural embellishments as the column and
the capital. It is a matter of observation that the introduction of
antique ornamental features goes with a decline in constructive
power ; at the time of the Renaissance we have the remarkable
spectacle of a triumphant classical decoration, in great part
imposing its laws upon a system of construction which followed
quite different paths.
Before Christian church-building was openly permitted,
that is, before a. d. 313, western art, whether generally influenced
from Hellenistic sources, or particularly influenced by Rome,
76 THE SUCCESSION OF PERIODS
was in a fair way to accept the complete supremacy of the vaulted
construction originating in Mesopotamia and Iran. Had the
Western Church yielded like the Eastern, Europe would not have
been arrested in its architectural development during a period
of more than five hundred years. The collapse of western
architecture resulted from the readoption of the wooden roof
in place of the vault. If in this matter the Church had not
followed the Temple, northern art would have been spared
a wrong turning. The S. Peter's of the fourth century ou^ht
by rights to have inherited the style of the basilica of Constantme
in the Roman forum ; it should not have been necessary for
Europe to wait until the sixteenth century to pick up the dropped
thread under quite different conditions. What struggles and
efforts fill the interval between the erection of the old S. Peter's
and that of the new, before the Roman vaulted basilica, trans-
formed in the Armenian sense, could continue its progress in
the West !
"^ The basilican architecture of the West may thus be divided
into the following periods : the period of the wooden roof ;
\ that in which the eastern vaulted building was transmitted to the
■ South ; finally, that of independent development in the North,
' lasting until the domed building attained a compromise with
the long-naved plan, and made its reappearance from the South,
anticipating any northern attempt in the same direction.^ It may
be that the future will solve the problem by discovering that the
external buttressing which we associate with Gothic and the
central-domed system of Armenia were created one for the other.
During the Early Christian period proper, the West was the
passive recipient of types transmitted from the Mediterranean
and the East. There can be no question of national styles in
the western church-building of the fourth century. The common
Hellenistic type of long building with wooden roof was predomi-
nant, and became the chosen form of the Roman Church. It may
be called the fatality of western architecture that neither Con-
stantine nor his successors built either of the two great martyria
of S. Peter or S. Paul with vaulted roofs, or in the form of the
domed basilica. We may suppose that the stream of oriental
influence ceased in Rome when the building of Constantinople
was undertaken, and the national architecture of the Armenians
and the Mesopotamian Syrians began ; while the groined vault,
which had become a characteristically Roman feature, was no
IN WESTERN ARCHITECTURE 77
longer in demand in the construction of columned basilicas.
Barrel vault and dome, the two essentials of church-building in
the East, could not permanently establish themselves in Christian
Rome. Yet the former had been transplanted to the western
capital by architects like Apollodorus of Damascus, who had
given them expression on the grandest scale in the Temple of
Venus at Rome, and in the great Baths and Fora, all examples
of vaulted construction in brick.
One reason why the timbered roof was so hurriedly adopted
in Christian church-building may perhaps be sought in familiarity
of the people with the type of the pagan temple. In the Iranian
region such models only occur in isolated examples like the
Temple of Garni in Armenia ; there was thus no general incentive
to reproduce the model. A cause which might have been expected
to dissuade Rome and the Hellenistic countries from adopting
it was its inflammable nature ; the Christians wanted not a shrine
or temple, but a building to contain a congregation, and the
assemblage of large numbers under a roof of this kind was
dangerous. It seems the fact that the timber-roofed basilica
only established itself in permanence where the antique temple
was widely represented. The regions which first followed
Persia in employing the vault were not those in immediate
relations with the Mediterranean area and its essentially Graeco-
Roman culture, but, on the contrary, those connected through
an active commerce with the main sources of vaulted architecture
in the East. The decisive influence in dissemination seems,
however, to have been that exercised by the mass migration of
the Goths westwards from the Black Sea. This people and the
craftsmen who went with them built vaulted structures in
groups where hitherto there had been only single, if conspicuous,
examples, as at Milan, designed to meet special needs of the Court.
In the Mediterranean area the retention of the timber roof
involved that of another classical feature — the column. This
must be regarded as a legacy even more momentous for Christian
architecture than the wooden roof. For while the combustible
roof was ultimately displaced through fears for the safety of the
congregation, and through the pious wish to build for eternity,
the column was never superseded.
The safer vaulted construction of the East had gradually
to force its way against old prejudice favouring the type of the
Greek temple. Progress was necessarily slow ; a process reaching
78 THE SUCCESSION OF PERIODS
its end in Persia as early as the fourth century, required more
than a thousand years to struggle to victory in the West against
the combined influence of Hellenism and Rome. And the victory
was never complete ; even to-day the North has to fight for its
independence against the Hellenistic Southern style which
triumphed so long ago in Europe. It was an event of deep
significance when Christian architecture in the Mediterranean
area transferred to the church, now required for congregational
use, features from the pagan shrine intended to house the god's
statue, a type of building derived from wooden construction ; it
was an event of no less moment when in contemporary Armenia
a monumental type arose which in its turn failed to meet the need
of the later Church for length in its places of worship. Between
these two extremes, development verged from one side to the
other until the West began to cover its churches with barrel vaults
and multiplied domes, and then entered upon its proper northern
path, while the Orthodox Church of Eastern Europe adopted
the domed Greek-Cross plan as its consecrated form. In the
East the centralized domed type triumphed over the long
church. In the West the opposite was the case ; the long axis
was retained, and in this the timbered roof and column were
more potent as determining factors than the eastern position of
the altar adopted from Oriental sources. For the altar existed
before orientation : even in the Syrian coast region the apse was
originally at the west end. In later times orientation, which
began in Armenia and Asia Minor, completely triumphed in
the West together with the long nave. Even to-day very little
is known of all these struggles for survival, the discovery of which
was first made possible by study of the ruined churches on the
eastern shores of the Mediterranean and in Armenia. In tracing
the destinies of Christian art in the West, I begin with that
region which, like Armenia in the East, was originally least
affected by Southern influence.
I. The tvooden churches of the North. The first question
which we have to ask is whether the North was capable of itself
arriving at vaulted buildings for congregational use. We may
be permitted to begin by comparison with China and India.
None of the three national religions of China required a hall to
contain a congregation. There was, therefore, no incentive to
build closed halls on a great scale ; and since local beliefs did not
demand buildings designed to last for ever, architecture naturally
IN WESTERN ARCHITECTURE 79
failed to advance beyond the stage of construction in wood. It
was otherwise in India, where wooden architecture predominated
among the Aryans down to the time of Asoka, but gave place
to a more durable form of construction when the idea of eternity
prevailed, at any rate in the religious field. A permanent
character was uniformly given to the constructed and free-
standing stupa with the stone rails enclosing the processional
path. More than this, the interiors of temples became permanent,
if only after the old indigenous fashion, so prejudicial to all
organic development, by which space was enclosed not by the
construction of walls and roofs but by excavation from the living
rock. But in India wooden building still survived in the monas-
teries, in palaces, and dwelling-houses.
In the North the wooden house and hall of pagan times
do not seem to have provided, as might have appeared obvious,
the model for the Cihristian place of worship. Nature herself
had from of old provided it under the open sky. Places for
religious assemblage, roofed in Hke house and hall, only came
in a short time before the admission of Christianity, and chiefly
as a result of its introduction. I therefore doubt whether it
is correct to assume from remains of temples in this part of the
world, that the temple really served as the point of departure
for church construction. It is hardly possible as yet to be
categorical ; but the North probably approached religious
architecture along the path already trodden by Greece, though
with the difference that the determining factor was not the
imitation of a shrine to enclose a statue, but the production of
an interior large enough to contain a congregation. This had
a remarkable consequence for the whole North, Celtic and
German and Slav alike ; what was wanted to meet the needs of
people living in small settlements far removed from each other
was not a ^w great buildings but a multitude of small ones.
The nature of the material employed counted for much ; but
a social or economic condition, the absence of cities, also had its
own significance.
The character of indigenous church-building in the North
is clearly reflected in the surviving wooden churches of Scandi-
navia. The student of Christian art is ill advised if he fails to
begin his examination of Northern architectural remains by
investigating these structures. In the North their position is
analogous to that of local types in Iran, those ancient develop-
8o THE SUCCESSION OF PERIODS
ments of Aryan wood-construction which formed the point of
departure in my discussion of Armenian churches. Many facts
of high evolutional significance have been overlooked or contested
through an incapacity to distinguish between the constraint laid
upon art by the religious movement from the South and the
creative freedom of architecture in the North, at any rate until
the immigrant southern influence became too strong to be ignored.
It is time that we began to take the large group of northern
wooden churches into account. There were conflicts and com-
promises with the South which might well explain many obscure
points both in the period of Oriental influence (Romanesque) and
during the growth of the crowning (Gothic) style in northern
church-building.
The wooden churches have more than one type. There is
the simplest house-form with four walls and gabled roof, exempli-
fied by the single-naved church at Hemse ; there is a type of
square plan, the so-called Valdres type (Fig. 14, opposite),
in which originally four masts seem to have been used in the
interior. I proceed on the contrary theory to that of Dietrich-
son ^ ; instead of regarding the wooden churches as copies of
models in stone, and beginning with those which have numerous
columns, I am inclined to construct their pedigree by placing
the four-columned Valdres type at the beginning, and tracing
the descent down to the twelve-columned type of Borgund,
where the church was built about a. d. 1150 and received its
present form in A. d. 1360 (Fig. 15, opposite). Here the
four single masts are replaced by groups of three, the result
giving that total of twelve columns characteristic of Aryan and
Islamic wooden architecture from India to Spain. At Borgund
we already find the long nave which, in satisfaction of liturgical
requirements, was afterwards the condition of all further develop-
ment. In any case, the introduction of masts in the northern
wooden churches may have met half way the forward move-
ment of the basilica from the South (Fig. 16, opposite).
There must, moreover, have been in the northern system of
wood-building some cause tending to the supersession of the
barrel vault by the groined variety ; this cause is perhaps to be
found in the method of supporting the clerestory and roof on
spaced interior supports instead of upon the outer walls.
The wooden architecture of the North may ultimately
* Die Holzbaukunst Norwegens, p. 17.
14 Valdres, typical wooden
church ; plan ; after Dietrich-
sen. See p. 80.
15 Borgund, wooden
church ; plan ; after
Dietrichsen. See pp.
80, 117.
17 Triforium of wooden
churches ; after Dietrichsen.
See p. 96.
16 Borgund, wooden churcli ; interior ; after Dietrichsen.
See pp. 80,93,96, 117.
IN WESTERN ARCHITECTURE 8i
derive from the same influences as the wooden churches of the
Ukrame, the stone reproductions of wooden dwelHngs in the
temples of Kashmir and a whole group of Indian buildings
with domes built by continuous corbelling. I will take only
a single piece of evidence, the so-called cushion capital of
Romanesque art. It is a mistake to suppose that the Lombards
invented the type and introduced it into Germany. The truth
probably is that it was invented in Armenia and carried west-
wards into northern Italy. In the Ararat region of Armenia it
is so densely represented that we can have little hesitation in
describing it as an indigenous form. Yet it might be precipitate
to explain its appearance in Germany directly by importation
from Lombardy or Armenia. Most likely it appeared in Armenian
and m German art from a common source, the wooden architec-
ture of the Aryans. It must always be remembered that the North
is every whit as cardinal in the movement of art as the Mediter-
ranean area in the South. We cannot therefore conclude just
because certain forms appear in Armenia, Lombardy, and
Germany, that one of these countries inspired the others, much
less that all were inspired by the South ; quite apart from their
mutual relations, all alike may have produced an identical form t
through mere technical necessity, for all had to deal alike with V
timber, the usual raw material of the North. We shall only
begin the proper study of this wooden architecture when we at
last bring ourselves to recognize its fundamental value both for
the northern area itself and for the border regions towards the
south, such as Armenia and Lombardy. It is to be hoped that
the discoveries at Oseberg may point the way in this direction.
II. The Oriental Art of Western Europe— Romanesque
Just as the dome on a square plan spread from Iran into
Armenia, so by ways at present not fully known the barrel-vaulted
church came into Gaul from Mesopotamia and the interior of
Asia Minor, there to undergo a development no less logical
than that of the dome in Armenia. Exploration and research
both in East and West tends to bring the Goths into the fore-
ground as the disseminators of this vaulted style, that Teutonic
tribe which taught the North its runes and once migrated, probably
from Sweden, across East Germany to the shores of the Black
Sea. Let us recall a single character in Gothic history, Ulfilas,
»45i G
82 THE SUCCESSION OF PERIODS
the translator of the Bible, who came from Cappadocia, and
accompanied the Goths across the Danube. But the tribe
pressed forward even farther ; its paths led it into Greece, Italy,
Gaul, and Spain, Wherever the Goths settled we find traces of
their art on an extensive scale, though both Ravenna and Rome
form exceptions. For though individual martyria, such as the
tomb of Theodoric, are domed, yet in Ravenna the church for
general assembly remained the timber-roofed basilica with long
nave. It is true that Ursus appears to have erected a vaulted
basilica, but, if so, it had no successor. The case was very different
in Milan. There Ambrose himself must have resorted to vaulted
construction, though S. Lorenzo alone survives to attest the
influence which he introduced. We shall return to this subject
below.
It was the Visigoths who laid in the West those foundations
on which vaulted construction grew up with the help of ever
fresh inspiration from the East. It may be true that the timber-
roofed basilica predominated in Gaul through Greek influence,
and among the Franks through Roman ; nevertheless the manus
Gothica did not cease from its activities. In Spain continually,
in France exceptionally, there arose buildings of ashlar which
the texts repeatedly ascribe to Visigothic hands. The relations
between the Goths and the East become convincing when
we contemplate the Tomb of Theodoric at Ravenna, unique
with its massive ashlar. The tombs of the Armenian kings in
the citadel of Ani on the upper course of the Euphrates must
have looked like this ; and it is related of the tomb of Sana-
truk (between a.d. 75-110) that the Persians, when they tried
to rob it in a.d. 350, were prevented from gaining entrance
by its skilful construction and the careful joining of its massive
masonry. The Armenians and the Goths were in contact with
each other in Asia Minor and on the Black Sea. Ulfilas came
from the Armenian borderland in Asia Minor; the language of his
translation of the Bible shows many traces of Armenian influence.
While whole Teutonic peoples migrated into Western
Europe from the East, there was a movement in the contrary
direction not to be overlooked when we are considering the
possibility that artistic forms travelled westward from the Black
Sea and the Mediterranean. Just as in our own time the whole
Mohammedan world would make pilgrimage to Mecca, so in
the first millennium of our era every Christian girded up his
IN WESTERN ARCHITECTURE 83
loins to visit the holy places of Jerusalem. The Crusades are
only intelligible as the end of a great movement which had
a stronger mastery over the minds of men than we are apt to
suppose. These continual pilgrimages, the itineraries of which
were carefully mapped out, led at the beginning of the period
into what was still the Promised Land of art. The sights which
met the pilgrims' eyes were not lost to memory without leaving
a trace ; on their way they saw the buildings of Armenian
Cilicia, at the end of the journey the venerable monuments of
Jerusalem and Bethlehem. We are just beginning to realize the
possibility that the Church of the Holy Sepulchre was not simply
a timber-roofed basilica with galleries, but a church with a dome
over the crossing.
If this was so, the surprisingly early appearance of this
type in Europe finds a ready explanation ; it cannot be questioned
that Armenians were quite capable of constructing it in the time
of Constantine.
The monasteries were also important agents in promoting
a lively intercourse between East and West ; other contributory
causes were the traditional poHtical relations between Rome and
Byzantium, and the continued interchange of views between the
Churches in countries round the Mediterranean. Monastic
institutions spread from Egypt, Syria, and Asia Minor, and every
new foundation meant an intellectual and artistic influence.
Thus no less a man than Cassiodorus adopted for his newly-
established monastery, Vivarium in Southern Italy, the system
of theological study ^ devised by the schools of Nisibis, and by
this action ensured its transmission to the West : remains of his
foundation point to the adoption at the same time of vaulted
construction. Marseilles and its neighbourhood formed the
terminus of a sea-route which very early rivalled in importance
the land-route followed by the Goths as a means for the penetra-
tion of the West by Oriental influence ; from the seaports this
route was continued in the interior up the Rhone and into the
valley of the Rhine. It was not used by the Church alone ;
it was above all a trade-route. It brought a swarm of Syrian
traders, in whose train doubtless came many Armenians. In
the preceding chapter we made acquaintance with a number of
individual buildings which confirm the presence of Orientals in
Western Europe, notably S. Germigny-des-Pres, which, so far
1 The Trivium and Quadrivium.
G 2
84 THE SUCCESSION OF PERIODS
as its type goes, might almost as well stand on Armenian soil,
and in its decoration, both of stucco and mosaic, betrays no less
clearly its connexion with the Iranian area. Another piece of
evidence which points to Eastern influence is the statement that
at this period Charlemagne sought the collaboration of Greeks
and Syrians for the emendation of the Gospel text, a statement
confirmed by the character of various miniatures in Carolingian
manuscripts. There was thus no lack of intermediaries between
the distant East and those West Prankish and Rhenish territories
which played the leading part in the earlier Middle Ages. Nor
were the above routes the only means of communication. Much
reached the West not directly by land or sea, but indirectly
through Upper Italy. From the fourth and fifth centuries, when
they were capitals of West-Roman Courts, Milan and Ravenna
had been centres of an intellectual and artistic life not uncon-
nected with the prevalent Oriental influences. In the first three
Christian centuries Rome had derived its artistic vigour from
Alexandria ; with the fourth century she continued to lose
ground. But in proportion as Rome declined, Milan rose and
pursued paths of its own. The landmark of these times,
S. Lorenzo, points to the region in which the new art had its
roots. As in liturgy and church music, so in architecture, the
initiative lay not with Rome but with Hither Asia. We have
already seen that the plan of S. Lorenzo, the domed square with
niche-buttresses, is the primary unit in Armenian church-building.
And reproduction of this kind could not have stopped here ;
S. Lorenzo may be exceptional ; but vaulting in national
church architecture must have been widely adopted at an early
date. The relations of S. Ambrose with the spiritual leaders of
the Greek Church in inner Asia Minor may well have embraced
such matters as church-building and decoration, for the letters
of the Fathers show beyond all doubt that these were subjects
of correspondence at this time. Like Bishop Ursus of Ravenna,
who before a. d. 400 built his vaulted Basilica Ursiana, S. Am-
brose may have himself erected basilicas with barrel vaults. The
decoration of the old cathedral of Ravenna, unfortunately replaced
by a more modem church, is happily known to us through
the description of Agnellus, which proves that the Iranian
method of covering walls in the north of Italy, to which we shall
recur, was usual. The nature of the decoration is described
by the letters of Nilus of Sinai and Paulinus of Nola : the
IN WESTERN ARCHITECTURE 85
nave walls were ornamented with hunting and fishing scenes
and figures of beasts ; the apse had a landscape. For Agnellus,
the ninth-century writer, these things had become an enigma ; but
there are still many traces in Ravenna of this more ancient phase.
It may be that in these early days the ground was prepared
for the commanding and original position assumed at a later time
by North Italy in the development of Western architecture, in so
far as this part of the country became a centre of distribution for
the Eastern vaulted style. To this various causes would naturally
contribute ; there were the direct relations between Milan and
Asia Minor, and those between Ravenna and Antioch ; above
all, there was the presence of Armenians among the immigrant
Goths. These Orientals were more intelligible than were the
Romans to the Lombards, who on their arrival were still in the
stage of building in wood ; they became indispensable when
the invaders adopted a new method of construction. I will
say nothing here of the hypothesis that the artificers known as
magistri commacini may have been closely connected with these
developments ; it is much more important that actual features
in Lombardic building point to Armenia.
Among such features we may especially note the Iranian
squinch, frequently used to effect the transition from the square
ground-plan to the circular plan of the dome, and the employment
of square tower-like drums with windows. But Armenian
influence appears also in long-naved churches ; it may, for
Instance, be detected in the so-called cushion-capital. In Italy
this form seems first to occur on a stone sarcophagus at Lambrate ;
in this connexion we may recall the tradition, unauthenticated
though it is, that Daniel, the favourite stone-mason of Theodoric,
was an Armenian. When we also remember that at a later period
no less than three Armenian cavalry regiments were quartered
in Ravenna, we shall see that the presumption of a national influence
is by no means impossible. It is, however, of more importance
that the cushion-capital became a characteristic mark of the
Lombard masters. It need not have come by a roundabout way
through Armenia ; but may have entered Italy directly from the
North through the influence of wooden construction. It was not
so with vaulting, which I shall now discuss in connexion with
individual types and varieties of church buildings in the West
itself.
86 THE SUCCESSION OF PERIODS
A. Barrel-vaulted churches.
The dominant kind of vaulting in the West was at first
not the domical but the ' barrel ' type. It is easy to see why :
the long church was already well established before there was
any idea of vaulting. But when vaulting was resumed in the
West, solutions of the problem how to roof long churches with
barrel vaults were ready to hand in Mesopotamia, in Asia Minor,
and, after the fifth century, in Armenia. In Iran itself another
method had been introduced, one based on the dome, in which
a long hall is covered by a series of cupolas in a line ; this we
shall find employed later in the south of France. After all this
there need be little cause for surprise when we find the West
treading in the footsteps of the East. I shall now discuss the
several provinces in which there was originally a distinctive
national character, concluding with a few words on the loss of
individuality by the general adoption of the northern Gothic
style.
The barrel-vaulted churches of the West have a nave only,
or a nave and two aisles, with or without rib-arches. These
features existed in the East in earlier times. But the vaulted
basilica, that is, a three-aisled vaulted building with a clerestory,
existed in the East only in isolated examples ; the same was the
case in the South ; the vaulted basilica was widely distributed
in the North alone. Should not this simple and fundamental
fact in itself lead to the conclusion that South and East go
together, but that the North follows its own path, starting from
wooden architecture and the timber-roofed basilica, and finding
its own northern solution in what we call Gothic ? I know that
a hundred details prompt to the rejection of this hypothesis,
but is it not advisable in our search for guiding principles at least
to keep it before us ? The vaulted church without clerestory
arrived in the West, so to speak, at a bound ; the vaulted basilica,
on the other hand, reached its final development only by degrees.
Does not this itself suggest the adoption of ready-made forms
on the one part, and a process of free experiment on the other ?
And is it mere chance that the adoption of the ready-made is
found in the South and the systematic experiment in the North ?
Did not the wooden churches of the North prepare the advent
of Gothic by means of their tall, mast-like pillars in their interiors,
which made clerestory hghting possible ?
IN WESTERN ARCHITECTURE 87
a. Barrel-vaulted Churches zvith single nave. The starting-
point here is the hall. The form is widely distributed in all the
Oriental regions in touch with Mesopotamia. Nothing was
known of this fact when the academic theory as to the barrel-vault
in the south of France was first formulated. We were told that
since antique remains- show it to have been the prevalent Roman
method of covering long rooms, the Roman method was the
obvious inspiration for the French architect, and thus we arrived
at ' Romanesque ' art. This theory is contradicted by recent
investigations in Armenia and Asia Minor and among the
Visigothic churches in Spain, proving that the Eastern barrel-
vault is not merely sporadic in the West, like the dome, but widely
distributed, with numerous monuments in groups. In Spain
a mass of important material has been discovered which my
opponents would fain treat as they have treated the material in
the East, refusing it significance for the early history of develop-
ment on the plea that it is of post- Mohammedan date. The horse-
shoe arch plays a special part in this discussion. This also is
a characteristically Armenian form, and may probably be traced
back to Aryan wood-construction. Its appearance in Spain at
the same time as the vault is significant. No amount of critical
gymnastics can permanently obscure the truth as to facts like
these.
In Mesopotamia barrel-vaulted single-naved buildings, with
or without rib-arches, are contemporary with the earliest Chris-
tian churches, and appeared in the second to third century ; they
may well have been indigenous in Armenia as well, and the fifth
century found them so numerously represented that examples
still survive. It need not therefore surprise us to find them
representing the most ancient element in the early church
buildings of France, especially in the south, in Aquitaine and
Provence, the heart of the Visigothic and Burgundian states.
They are long churches with barrel -vaults and rib-arches, the
nave leading up to a hemispherical apse. They often show
Armenian characteristics ; they have polygonal apses and very
massive walls ; above all, they reject the wooden roof, the tiles
being laid directly on the extrados of the vaults ; we may add
to the list their tendency to end the chancel with a trefoil. It
is true that surviving examples are not earlier than the year
A.D. 1000 ; but if once we grasp the possibility of a connexion
with the Visigoths and their Armenian architects, fresh research
88 THE SUCCESSION OF PERIODS
may well bring earlier remains to light. Sometimes the pier-
projection is replaced by that conspicuously I rano- Armenian
motive the double pilaster.
Spain has also a whole series of one-naved buildings with
barrel-vaults on rib-arches, and actually dating from the Visi-
gothic period. The chief examples are the church of the Virgin
at Naranco, a structure recalling the hall of the Tak-Ivan in the
south of Persia ; the church of Santa Cristina at Lena, and other
buildings. Here we seem to have contemporary evidence which
may be brought into direct connexion with the Oriental influence
introduced by the Goths.
It is a fact not without importance that we find in the
south of France a tentative movement towards the Armenian
' domed hall '. The vault is carried not by the walls, but on
projecting piers in front of them, the walls becoming thinner,
and the piers dividing the interior into shallow bays. As in
Armenia, we here find buildings in which the treatment of space
is handled in a masterly manner for which it would be hard to
find a parallel. The chief example is the cathedral of Orange.
b. Hall-churches. In the case of the so-called Hall-churches,
with three aisles but no clerestory, it is again commonly assumed
that Roman buildings, such as the Bains de Diane at Nimes,
served as models. But we must observe that the hall-church
is the common form of the three-aisled long church in Armenia
and Asia Minor. Here again the old dogma must be retested
in the light of new facts. The Roman barrel-vault is always
supported by the walls, never, except perhaps in cisterns, on
piers or columns. The most individual attempt to support the
arch carrying the vaulted roof on a pier projecting from the wall
was first made on a large scale in the east of Syria ; the earliest
three-aisled churches with such piers may have originated in
the interior of Asia Minor ; in Armenia they began in the fifth
century.
With the hall-churches we reach a type which had a wider
area of distribution. They spread from Lombardy and the
Rhone valley, in the south into Spain, in the north along the
Rhine into Westphalia. They also flourished in Auvergne,
which introduced a hall-church with galleries, just as it adopted
the galleried domed-basilica in opposition to the Armenian type
of one-domed church. In its southern region the West begins
with the hall-church, a logical development similar to that
IN WESTERN ARCHITECTURE 89
begun by the North with the basilica. I shall not here treat this
subject in detail, but select only such points as properly concern
the present volume.
B. Churches with several domes.
Allusion was made in the last chapter to a group of long
churches in France apparently connected with similar buildings
in Iran. We should no more regard these as of French
creation than we should attribute, for example, the early church
remains on Bulgarian soil to Bulgarian genius. Like the type
of the barrel-vaulted churches with single nave, this type also
is an importation, more ancient than the Frankish Empire, and
probably of Eastern origin. These churches occur in the same
regions as those with barrel-vaulted halls, and chiefly in Aqui-
taine ; in Perigord they completely superseded the barrel-
vaulted type. Armenia, as we have seen, was not their place of
origin ; they came from Iran, their route, so far as it can be
traced, leading from the region of the Nestorian churches and
of Aleppo to Constantinople. The employment of the pointed
arch is additional evidence of this. It is remarkable that in the
French churches the domes do not rise from squinches, but
from spherical pendentives, though these are not constructed
with wedge-shaped stones but, as in Armenia, with overlapping
horizontal layers. The blind arcading in the interiors also
points to Armenia. The chief group shows several such domes
following the long axis of the church, but in the case of S. Front
at Perigueux their cruciform arrangement resembles that of
the lost church of the Holy Apostles at Constantinople and of
the wooden churches in the Ukraine. There has been much
controversy whether these French churches are or are not
Byzantine. The answer is certainly in the negative. But they
afford perhaps the best examples of a movement from Iran to
Europe which did, indeed, pass by way of Constantinople, but
died out in a region much farther to the west, where it found
its chief representation. Like the Armenian domed churches,
the churches of Perigord are distinguished by one common
characteristic. Their primary purpose is to enclose space ; the
architects dispensed with decoration in the achievement of the
desired effect ; it is from their simplicity that these buildings
derive their greatness. S. Mark's at Venice is a milestone upon
90 THE SUCCESSION OF PERIODS
their way from East to West, though here the splendour of
decoration shows an intermediate influence, coming not from
any area alHed in feeUng to Armenia, but from the Byzantine
court. It was therefore an error to connect S. Front primarily
with S. Mark's, even a greater error than to hold the old belief
that the cathedral of Aix-la-Chapelle was no more than a copy
of S. Vitale at Ravenna. For the eff^ect of its interior S. Front
should be compared with the cathedral of Ani, or with Thalish ;
so also should Notre-Dame at Le Puy. When this comparison
is made it will be seen how far East-Aryan conception in its
impressive unity excels the West-Aryan agglomeration of spatial
units.
In their attempts to discover the origin of these single-naved
domed churches in France, writers have often brought the
Crusades into the picture, and have drawn special attention to
similar buildings in Cyprus, altogether overlooking Armenia, and
Cilicia with its Armenian art. More than once these discussions
have also brought into notice the almost contemporary movement
connected with the Manichaeans, Paulicians, and Bogumils,
which was ended by the crusade against the Albigenses. One
imagines oneself transported to Armenia when one reads of the
rejection of the Eucharist and remembers what struggles with
the Greek Church centred in this point in the countries where
this type originated. Much of the plainness in the decoration of
these buildings, such as simple profiles and blind arcading, might
be explained on the hypothesis of such a connexion. Evidence
of certain Persian features is also not without significance.
C. The groined vault.
The groups hitherto discussed possess only the barrel-vault
and dome. But developed Romanesque employed the groined
vault. Must this be ascribed to the revival on Frankish soil of
the Roman moulded variety, despite the fact that the South
transmitted the timber-roofed basilica, and only the East the
vaulted building ? This is the popular hypothesis. In my
judgement the capacity for revivals, by which I mean deliberate
restorations of Southern forms in the North, was lost for centuries,
and in fact presupposes a very different intellectual plane to
that attained in the early years of Romanesque architecture.
Most people will perceive an underlying contradiction here.
I
I
1 8 Salah, Church of S. James, type with
transverse nave ; vaulting of nave ; photo.,
G. L. Bell. See p. 91-
19 Ereruk ; South view. See figs. 8, 9
and p. 69.
20 Sofia, Church of S. Sophia, cruciform domed church with
three aisles ; plan; after Filow. Seep. 91.
IN WESTERN ARCHITECTURE 91
The earlier suggestion of a Byzantine influence was nearer the
mark ; it was at least so far correct that it derived vaulted con-
struction from an Eastern source. This construction may have
contained the germ of the groined vault subsequently developed
by Aryan logic. The Greeks began with a wooden house, the
Armenians with a walled square surmounted by a dome. What
did the West- and East-Aryans of the South make of these
elements .'' What did the North make of the groined vault ?
The Goths brought to the Northern Aryans, among other
things, a kind of vault well known to us in Mesopotamia (Fig. 18,
opposite), and traced on its western migration as far as the
church of S. Sophia in the capital of modern Bulgaria ; this was
a barrel-vault in which the bricks are so laid as to suggest groining.
For this reason the old cathedral at Sofia has been erroneously
described as a Romanesque church with groined vaulting, and
only on the strength of this supposition was it possible to dispute
its age. Actually, it is a fine three-aisled domed church with
trefoil end, and the cathedral of Thalin is hardly an older
example. The dome-like roof is concealed within a square
tower such as that of the tomb of Galla Placidia at Ravenna,
that at Tekor in Armenia, and that at Khakh in Mesopotamia ;
such a tower as that above the crossing in the church of the Holy
Sepulchre at Jerusalem, in the church of the Saviour at Spoleto,
and in many post-Carolingian churches in the West, especially
in the region of the North Sea and the Baltic from the British Isles
to Sweden.
The barrel-vaults at Sofia, which suggest groining, and are
connected with the square groin-vaulted tower, may have
travelled in the same way, though at present there is no tangible
evidence. One thing only is certain, that alike in the Mesopo-
tamian group and in the isolated example at Sofia (Fig. 20,
opposite) the barrel-vault is first narrowed by building up
the sides ; thus small squares are formed between the rib-arches,
and these are covered from all four sides by courses of bricks
parallel to the sides of the square. In this way four triangles
are formed, the sides of which meet on the diagonals of each
single square and unite in a small terminal square at the top.
Now the simple groined vault of the West shows the same kind
of construction at the springing of the vault, carried out by means
of horizontal courses, each projecting beyond that beneath it ;
it shows also the diagonal arches, that is to say, the triangles
I
92 THE SUCCESSION OF PERIODS
simply meeting in a joint. These arches do not project in the
single and unbroken longitudinal vault, but come into being
when this is met by transverse barrel-vaults of the same kind.
A distinctive mark of this work is seen in the fact that the direc-
tion of the joints is no longer vertical, but parallel to the axes of
the several triangles.
In the north of France and in Germany the groined vault
is the starting-point of a logical development at the same moment
(about A.D. iioo) when we may note tendencies towards a
Proto-renaissance in the South. Surviving remains of antique
architecture were here the sources of inspiration (Cluny, Autun).
The old theory is now disputed that this movement had as
a direct consequence the introduction of the sculptured human
figure into Western architecture ; to this point I shall return
below. However, an end was made of this incipient influence
by a bold action on the part of the North which rid Western
Europe for centuries of this retrograde Southern movement.
Would that it had stemmed it for all time !
The striking and essential difference between the Oriental
influence in France and Germany deserves careful attention.
In contrast to the gay variety in France with its wonderful wealth
of local difference, we find in Germany a purposeful effort after
the uniform development of all available artistic values, only
comparable to the similar activity in Armenia. A fixed type
resulted in both countries, a long church, which in Germany
had groined vaults, in Armenia dome and barrel-vault. Very
striking are the resemblances in the mode of decoration, the
well-defined unity of the interior, and the accentuation in the
exterior of a clearly conceived design. Probably the genius of
the North succeeded in asserting itself more fully in Armenia
and Germany because these countries were less led astray by
influences from the South, especially that of classical art. But
what may be fairly described in each case as the national employ-
ment of the cushion-capital suggests other possibilities. In
Carolingian times there must still have been a northern route
between Armenia and the Rhine. Architectural features at
Wiirzburg, in Bohemia, and Cracow attest the migration of
Armenian forms such as the sexfoil,the conched square, and the
quatrefoil. In this connexion we may note the popularity of
the trefoil in Rhenish cathedrals. A comparison of the eleventh-
century S. Maria in the Capitol at Cologne with the seventh-
IN WESTERN ARCHITECTURE 93
century Cathedral of Thalin awakens a lively desire for further
investigation in this field. Further, the conformation of the old
cathedral of Cologne and the persistent tendency to such rich
combinations in Rhenish cathedrals both find their explanation
in Armenia, especially in individual features, such as the dome
over the crossing and in the superimposed turrets for bells. In
France a single and logical development first came in from the
North when the old forms adopted from the East had ceased to
be effective.
III. The Northern Art of Europe : Gothic
What we choose to call Romanesque in Europe is based
upon a Southern form of church, the timber-roofed basilica,
translated into terms of oriental vaulted construction. Gothic
represents the Northern version of this art. The unfamiliar
phrases. Eastern art of Christianity in the West and Northern art
of Christianity in the West, will at first seem hard to accept ;
but they fit the facts, and we shall soon grow used to them.
The Gothic treatment of the long vaulted church resembled
that of the Iranian dome in Armenia. The dome rested first
on the walls, then on free-standing piers within the walls ;
finally wall and pier were linked by narrow arches, and a system
of central supports created. This kind of abutment was brilliantly
employed in S. Sophia at Constantinople. Like the Armenian
domed hall, it is * Gothic ', though with interior abutment and
plain external walls.
The connexion with the foregoing orientally inspired
development is obvious. But I must insist with emphasis on
the common mistake of ignoring the Scandinavian wooden
churches. These churches, in complete independence of the
South, created their own church type in wood, a * basilican ' type,
and therefore without vaulting (Fig. 16, facing p. 80). By their
vertical development and their treatment of space they show various
features which appear to be of some importance for the growth
of vaulted architecture in the North. If it is remembered that
at the time when the groin-vaulted basilica originated, the central
part of Europe must surely still have had a flourishing wooden
architecture such as survives even to-day in Scandinavia, here
and there in other regions, and of course in Russia, then we ought
to consider the Seine valley in the light of a nursery for these
94 THE SUCCESSION OF PERIODS
Northern developments no less than for those from other quarters.
We shall come back to this subject later. Here I draw attention
to a few features as conspicuous in this logically progressing art
of the North as in Armenia under similar conditions ; the im-
pression is thus conveyed that the development in both countries
was only rendered possible by the promptings which both received
from the Northern spirit. But there is also the possibility of an
Oriental influence on the growth of Northern art, the introduction
of which is usually ascribed to the Crusades. French scholars,
VioUet-le-Duc, de Vogiie, Courajod,Dieulafoy,and others followed
these tracks more than half a century ago, but much has onlybecome
intelligible through our new knowledge of Armenia. The fact
deserves special mention that in Armenia there are buildings
which Schnaase, Texier, Lynch, and others found inexplicable
except on the supposition that they had at least been recon-
structed or restored by a few ' Gothic ' architects from the west
of Europe. The chief example of this has always been the
cathedral of Ani (Figs, ii and 21-23, facing pp. 71, 94, and 95).
This church was begun before a.d. 989 by the great architect
Trdat, and completed in a.d. iooi. The use of the pointed
arch to support the dome and the absolutely western style of
the four piers with their clustered columns (Fig. 23, facing p. 95)
were the features which chiefly suggested the interposition of
a western hand. In reality the appearance of these very features
can be traced through their logical development during the period
between the seventh and tenth centuries, just as certainly as the
recessed Romanesque portal.
And, as a matter of fact, with the transformation of the old
pilgrimages into the Crusades there began a change in relations
which had hitherto been confined to the introduction of oriental
artistic forms into the West. There now began a counter-
migration. The distinction amounted to this, that the Oriental
forms set on foot a movement in the West very fruitful in develop-
ment, while the western forms in the East remained, at any rate
in the domain of architecture, alien importations, sterile and with-
out effect. This is well shown by the relations between Moham-
medan art and that of the West. Whereas we may with probability
discern the influence of this essentially Iranian art in certain
forms of abutment, in the use of the pointed arch and of the
pierced stucco window-slabs filled with coloured glass, we find
on the contrary no evidence that Armenia or Islam adopted any
I
21 Ani, cathedral ; interior from North-West.
-^w M t r r
22 Ani, cathedral ; plan. See pp. 72, 94.
23 Ani, cathedral ; bases of piers with clustered
columns. See pp. 72, 94, 95, 145.
24 Thalin, cathedral ; detail of South-West pier.
See pp. 95, 145, 146.
IN WESTERN ARCHITECTURE 95
western forms. The Mohammedans carried off as a trophy a
Gothic doorway from Akka and placed it in the Mosque of
Kalaun at Cairo ; at Jerusalem they adopted much that was of
western origin ; but they went no farther. It was the same in
Armenia, where the comradeship of the Crusaders and the
Rubenids of Cilicia afforded ample opportunity for exchange.
There can be no question of a penetration of Armenia by Gothic
forms ; an influence of Islamic and Armenian forms on the North
through the Crusades is more probable.
What the cushion-capital was for the openly Oriental influence
in Europe, a Leitmotiv, invaluable for comparative architectural
study in Armenia, Lombardy, and Germany, such, in the case
of the purely Northern influence, were the engaged shaft and the
pier with clustered shafts. The engaged shaft is a vertical
cylindrical member running up wall or pier, and having no
connexion with the free-standing column. This, indeed, can
only be clearly demonstrated in Armenia, where the column
was never admitted as a constructional feature, whereas the
engaged shaft had come in from the first with the Iranian
domed building. In Armenia the shaft appears both in the
interiors and on the exteriors of churches (Fig. 23, opposite),
in the latter case generally double, and associated with the
niche (Fig. 2, facing p. 62) and cushion-terminations at top
and bottom, but also at a very early date in the demarcation of
wall space within the church. But it was only in conjunction
with the pier that the shaft acquired its great importance for
comparison with the West. It was applied to the projecting
part of the pier as a half-cylinder, or, as three-quarter-cylinder,
placed in the comers between the projecting parts. The use
made of it is thus exactly the same as in the West at a later
period. In all this the shaft is not bound, like the column,
by any fixed relation between height and diameter ; it is an
unrestricted form, which in certain cases is indefinitely extended.
Another fact to be noted is that a kind of ribbed-vault was
developed in Armenia, the origin of which is to be found in the
porches placed in front of small churches when they were enlarged.
Fig. 25, facing p. 96, shows such vaulting from the monastery of
Aisasi, though it must be admitted that the example is a late one.
We see two massive ribs in the form of pointed arches which
embrace a small dome, and are buttressed by two quarter-arches.
There are other similar combinations, and it is possible that such
96 THE SUCCESSION OF PERIODS
stone structures had prototypes in wooden buildings. So much
for the features of importance for the comparison between
western and Armenian types. As regards the wooden buildings
indigenous to the North, I proceed to the following remarks.
It was suggested above (p. 80) that the wooden churches,
by inserting tall mast-like supports, may have helped to prepare
the way for the basilica advancing from the South. We may
conjecture that the development was from the simple type with
four columns to that with a group of twelve, and from this to
the unrestricted use in rows. This involved a calculated system
of meeting thrusts, which may have prepared the way for the
crowning achievement of northern stone-construction, the so-
called Gothic style. Observe how in the interior of the church
of Borgund (Fig. 16, facing p. 80), the masts are buttressed by
the frame-walls and the roofs of the aisles, and in Fig. 17, facing
p. 80, the detail of a triforium, where the method of buttressing
by the frame-walls is repeated in a somewhat clearer manner.
It seems to me that here the approximation to thrust-receiving
arches is no less clear than in Armenian or Romanesque buildings.
Other valuable observations can be made on the articulation of
the roof supports at Garde on Gothland. Such points as these
should be indefinitely multiplied as soon as our eye for these
matters grows sharper. My own country of Austria will certainly
yield valuable results.
Fig. 26, facing p. 97, a house in the Upper Austrian town of
Steyr, shows a characteristic example. It is just one among many
which represent the local type of citizen's house, and illustrates the
old indigenous wooden forms translated into stone. The inner
court of the house is surrounded by open galleries, which, how-
ever, are not visible on the ground floor to the left. In their
place we note piers before the walls, carrying long projecting
brackets which support the gallery with its pillars carved in
angular relief. This otherwise characteristic example unfortun-
ately lacks the usual pierced screens, sometimes wrought in
stone after the model of the old wooden prototypes. The
influence of church-building need not be denied. But the
essential point is that in spite of it the old domestic wooden
style is everywhere to be found, the style which may have had
far more influence than we think on the beginnings of church-
building itself. Unluckily it too succumbed to the artistic
ignorance of the humanistic movement behind Church and Court,
25 Aisasi, porch of monastery church ; Armenian ribbed vaulting.
See p. 95.
26 Steyr, court of a house ; photo., Reiffenstein. See p. 96.
IN WESTERN ARCHITECTURE
97
and Europe reverted to the habit of aping classical models.
Very few people at all grasp the fact that as a result of this rever-
sion, real understanding of the nature of art and its value to
life was in great part lost. In all fields of knowledge except
those of science and technology, the belief that the imported is
able to replace the indigenous has held its ground for centuries.
The consequence is that not formative art only, but the whole
of European life is under various aspects pretentious and insincere.
To resume the foregoing : the Christian art of the North means
at bottom the rebirth of a true European art, of a North-Aryan
influence in a Europe which after the close of the Hellenic period
had ceased to be independent and creative.
We have seen that northern art itself renounced its proper
character in so far as it conceded to the human figure in archi-
tecture a place only equalled for importance in India. A distinc-
tion must, however, be made. In India the suggestion came not
from the art of the immigrant peoples in the north of the country,
but from that of the older population in the south, just as farther
West it came from Egypt to the Greek art of Southern Europe.
In North European, or Gothic, art, however, the essential lies
not, as in India and Greece, in the human figure itself, but in
the draped figure — not in the body, but in the covering given to
the body by art. The figure itself is subordinated to form, as
in East Asia ; natural shapes become merely the vehicles of
rhythmical line. Moreover, these northern figures are in organic
unity with the body of the structure. The consciousness that
the various parts of the organism are thus naturally enlivened
leads, independently of the human figure, to a luxuriant over-
growth of vegetable and animal forms unequalled in any other
art, even in the South.
The North thus took the step from decoration to representa-
tion without fully subordinating the natural instinct of formative
art to exact reproduction of Nature. Representation gradually
triumphed, but was never carried to its logical extreme. That
was only done when styles of Italian inspiration established art
upon an intellectual basis.
IV. The Italian Mixed Style {Renaissance)
It was reserved for Italy to make Europe a second time
acquainted with the East-Aryan dome, Italy which continuously
maintained the closest relations with the Eastern coasts of the
2451 H
98 THE SUCCESSION OF PERIODS
Mediterranean. Aquitaine had introduced the multiple domes
of Iran ; it was for the Renaissance to recognize the essential
merit of the single-domed Armenian plan, and to give it a
permanent place in European architecture. The course of
development repeated that witnessed by the East in earlier times.
A Leonardo and a Bramante established the principle that the
dome is the centre of the whole. Then the Church, as patron,
won the upper hand and demanded a long building, and for
the second time in the history of Christian art Vignola discovered
in the ' domed hall ' the solution of the problem. But since in
the West the basilica was no less firmly established than the
church with central dome in Armenia, the dome was not able
to preserve the law of its being by maintaining its right position ;
it was advanced to the East end, where the old dome over the
crossing had stood. While, therefore, in Armenia the long nave
was imposed upon domical architecture, in Italy the dome was
placed upon the long nave. I will attempt a brief explanation
of this development.
By buttressing the dome as he did, Brunelleschi may be
said to have completed the Gothic cathedral of Florence in the
Armenian style. Looking at the East end from without, one
might take it for the work of an Armenian architect. The pointed
dome alone has a Western air, though the shell of the cupola
again suggests the East, and especially Islamic architecture. In
S. Lorenzo and S. Spirito, indeed, the affinities are rather with
Romanesque, the Oriental style of Western Europe ; and here
Brunelleschi adopted the dome over the crossing. But in the
Pazzi chapel he took a step which showed him in a fair way to
full understanding of the square as the right form for the bay
beneath the dome ; his plan for S. Maria degli Angeli in like
manner shows that he had in mind the Armenian eight-lobed type.
Alberti and Michelozzo, who had himself visited Cyprus, went
beyond Brunelleschi ; their ideas carry yet farther the possibilities
of the dome over a square plan. Yet Leonardo, as his drawings
show, was really the first who flung himself into the exploitation
of domical building, though at first he had no thought of adapting
it to churches ; a visit which he may have made to the Taurus
country would be enough to explain these Oriental designs. But
the architectural tasks entrusted to him at Milan and Pavia in
connexion with the cathedral, together with his activity in the
service of Francis I, directly confronted him at last with these
i..'\:i
27 Vagharshapat, Hripsimeh, a.d. 618, domed square with apse-buttresses
and angle-chambers ; South view.
28 Vagharshapat, Hripsimeh ; plan.
See p. 99.
29 Amman, square building with apse-buttresses ; blind
arcade with formal ornament. See pp. 123, 124.
w i t t i r
30 Ani, church of the Af)ostIes ; plan.
See p. 99
31 Korea, stone tomb ;
structure of the roof ; from
the Kokha. See pp. 122,
123, 124.
IN WESTERN ARCHITECTURE 99
problems as related to church construction. His influence in
any case seems to have brought Bramante to a point at which
he permanently abandoned the Gothic style and entered the
paths leading him to S. Peter's. In all probability Leonardo's later
sojourn in Rome only matured the plans first formed at Milan.
For Leonardo the central architectural idea was the octagon ;
but side by side with this appears time and again a plan corre-
sponding to that of the Armenian ' apse-buttressed square ' with
interior supports (p. 63). This type is, indeed, exemplified by
S. Germigny-des-Pres, and by S. Satiro at Milan, but both
are derivatives from Armenian models ; Leonardo may just as
easily have come into contact with the original type in Cilicia.
Could it be shown that he was also acquainted with the plain
apse-buttressed square without central supports, we should
have abundant proof that this was the case. Unfortunately,
Leonardo took the lamentable step of adorning the Armenian
structure, not in its proper style, but with classical features
after the ordinary fashion of the Early Renaissance. Significant
evidence of the constructional ideas which he possibly derived
directly from the East is afforded by the chateau of Chambord.
Here, after the manner of Armenian and Mohammedan palace
construction, he makes his dominant central point a domed hall
of cruciform plan with barrel-vaulted limbs, the very plan which
inspired Bramante's work at S. Peter's.
The history of this great church repeats as a solitary example
the destiny of Early Christian art in Armenia. As was the case
a thousand years earlier, a pioneer architect produced a monu-
mental design, placing the central dome, as in the Church of the
Apostles at Ani, above a quatrefoil (Fig. 30, opposite), but of
necessity interposing barrel-vaults between the dome and its niche-
abutments, because he disposed four chambers on the diagonals,
as in the case of Awan or the Hripsimeh (Fig. 28, facing p. 98).
Naturally, there is no comparison between the dimensions, but the
plan in itself is certainly similar. Bramante's design was not
originally intended for execution, and had to be modified in
essential features before it could actually be carried out. Yet
throughout it remained fundamentally Armenian, even when
handed over to Peruzzi and Michelangelo. In the end, the
demands of the Church for a long building were successfully
asserted ; the architects gave way as they had done long before
in Armenia ; and the result is seen in the S. Peter's of to-day.
H 2
100 THE SUCCESSION OF PERIODS
In Armenia greater foresight seems to have been shown ; the
main point was never lost si^ht of that the dome had to dominate
the whole interior, and be visible from the entrance in all its parts
right up to the crown.
Leonardo and Bramante more or less represent the archi-
tectural point of view of the Armenians in the fourth century,
though the dimensions of their buildings were very different.
The work of Vignola corresponds to the phase upon which
Armenian domical building entered after the fifth century under
the influence of the Church, its final achievement being the domed
hall. On these lines Vignola built his Gesu, the Jesuit church,
the type of which spread with the order over the whole of Europe,
and became the characteristic model for the Barocco style. The
difference between his solution of the problem and that of the
Armenians lies, as we have seen, in the fact that in Armenia
the long nave supervened on the centralized domical plan, while
in Italy the dome supervened upon the long nave. The results
are essentially alike, though in Armenia the dome had the last
word, in Italy the conception of the long church, the dome being
transferred from its proper position to the bay before the apse.
Bramante seems also to have manifested that feeling for pure
construction which we had occasion to praise in Armenian
buildings, in those erected in Aquitaine after the Armenian
manner, in the work of the severe German tradition, and in that
of early Northern (Gothic) art, the feeling which wholly sub-
ordinates ornament to construction and spatial effect. The
fashion of making fa9ades, external or internal, by the employ-
ment of antique elements had not yet triumphed ; it must have
found little favour with Bramante during his earlier career, if
only for the reason that he also aimed at combining unity of
interior space with an impressive treatment of the exterior as
an organic whole, producing its effect upon the spectator by the
equal influences of all its parts.
The victory of the Renaissance throughout Europe was deter-
mined by the supremacy of Court and Church which was every-
where followed by the Counter Reformation. The Roman forms
were tacitly accepted as alone suited to a policy based on power.
It is a favourite assertion that with the progress of develop-
ment after Roman times the contrasts between Asia and Europe
grew more and more profound. In the early Middle A^es in
Europe the opposite was true ; even in the sphere of formative art
IN WESTERN ARCHITECTURE loi
political rivalry demanded compromise, because the West never
wearied of learning the tricks of impressive display practised by
the Oriental courts.^ Let the question be asked, on the other
hand, which method had the final word in architectural decoration,
that in which spatial and constructional form were brought into
ever closer harmony, or that which adhered to the columnar style
of classical antiquity only abandoned by the West during one brief
interval. Here, too, the East had already completed in Early
Christian times that breach with Greece and Rome which a proper
regard for our own independence bids us open anew.
Now that we have evidence of an Early Christian vaulted
architecture in the East, we have the chance of studying the
historical development of art along these lines. The old method
was to proceed by basing everything on literary sources, and
relying for the rest on general history ; that is, by fitting the
monuments anyhow into a ready-made frame without the least
regard for possible gaps, which might disarrange the whole
picture ; let us hope that this procedure is done with once and
for all. Hitherto we have been engaged in writing mere history ;
now at last we can begin as specialists to apply the comparative
method, observing each work of art on its own merits and for the
essential values which make it what it is. So long as the most
important comparative elements were unknown, this method was
excluded, and it was therefore quite impossible to grapple with
questions of origin. When such questions were in fact approached
in the light of some theory like that of an imperial Roman or of
a Christian classic art, the attempt was bound to come to grief
through two causes, an inevitably one-sided point of view, and
a palpable ignorance of the monuments in question.
In the first half of this book, ending at this point, the familiar
old Christian buildings have been evoTutionally considered, and
the types set in their proper places in the order of development.-
In the second half, which begins with the next chapter, it will
be our business for the first time to demonstrate the existence of
the oldest branch of Christian art, till now to all intents and
purposes totally ignored. In the process the complete contrast
between the usually accepted doctrine and my own will become
absolutely clear. In the succeeding chapters I shall again confine
myself to the Early Christian period.
' Architecture in the West reacted to Eastern influences introduced by the
immigrant Goths.
Religions without Representational Art
WE are apt to conclude from our own example, and from
that of Ancient Egypt, Mesopotamia, Greece, and India,
that all religious art is naturally representational ; thus
we never reflect that the earliest Christian art cannot possibly
have had this character, a point to be discussed in the next chapter.
I need here only observe that while Hamites and Semites in the
Nile Valley and in Mesopotamia had representational art, the
immigrant Aryans were at first without the idea of copying
nature, and only began to render deity in human form under the
influence of the southern peoples. The same is true of India.
Here Arj^ans were first induced to represent by the Dravidian
and other indigenous populations, just as the Greeks had adopted
the older Mediterranean art, the affinities of which were with
ancient Egypt and Mesopotamia.
In my work Altai-Iran I sought to explain that this difference
in practice results from a fundamental opposition of North and
South in their attitude towards art, an opposition which may also
be awakened by radically different economic and social conditions.
Thus the nomadic herdsmen of the desert and the steppe base
their art upon principles quite distinct from those adopted by
the agricultural peoples in the valleys of the great southern rivers
— those veritable forcing houses of culture. In the present
volume I shall broadly define the manner in which this suggestion
may influence our idea of religious art. In the province of archi-
tecture, we found that progress along these lines was blocked if
we persisted in beginning with the Mediterranean area, with
Rome and the earliest sepulchral monuments, which have hitherto
cumbered the foreground of Christian art ; advance was only
possible if we started from the opposite side, from Asia and from
the North. We shall find things the same in the province of
formative art.
The characteristic ornament of early Teutonic antiquities,
RELIGIONS WITHOUT REPRESENTATIONAL ART 103
still surviving in the decoration of the Scandinavian wooden
churches, has peculiar features shared by the art of Islam. Neither
the ancient Teuton nor the Mohammedan made use of representa-
tion on any extensive scale. The investigator approaching the
problem from the Asiatic side perceives affinities between their
respective arts the moment he considers them from the angle
of Altai and Iran. He observes that the nomadic herdsmen
roaming between North-east and South-west, especially the Turks,
had no representational art, and that the Aryan nomads roaming
between North-west and South-east were in exactly the same
case. The artistic motives which we find Teutonic peoples
bringing with them across Europe at the time of the Great Migra-
tions represent only the last phase of a very old development
gradually becoming clear to us, a free development flatly opposed
to that representational system of the southern peoples which
has alone attracted notice in historical times. There is no
question here of degrees of merit, as, for instance, between
a mature and an immature art ; the distinction is one of kind,
and concerns two different entities. Of these only one, the
representational, has counted in the eyes of European scholars
trained in the classical school of Greece and Rome, while these
same scholars, even those of the North, have never really
approached the other, though in the very nature of things it
should have been their first concern.
When one hears the dictum that decoration in a single plane
proves arrested development, while the plastic treatment of Greek
art proves maturity, one can but say that such a statement
betrays a fundamental ignorance of the distinction between
representational and non-representational art. It is of the essence
of non-representational art to decorate and fill a flat surface.
The idea of filling implies another idea, that of framing. Repre-
sentational art does not at first enclose its subjects ; it covers
the surface without order, or obliquely, or in superposed zones,
and at first itself begins by keeping to one plane. Non-repre-
sentational art on the other hand, with an enclosed space to fill,
never abandons the flat surface ; it is the human figure which
forsakes the single plane as soon as it is framed. This dis-
tinction is fundamental as regards the two methods. Non-
representational art was born out of handicraft ; it was from
handicraft that art sprang into being in the North and among
the nomadic shepherd peoples ; Gottfried Semper should have
I04 RELIGIONS WITHOUT
taken up his parable at this point. The origin of Southern art,
with its representational ideal, is on the contrary to be sought
in the subject ; from the beginning it pursued the imitation of
nature. Non-representational art ignores alike reproduction
of nature and disposition of forms in cubic space.
The conception which has hitherto prevailed could only
have originated, one might think, in that partition of art-history
which is concerned above all with the Mediterranean. The
student using the comparative method cannot be content with
it. His eye is directed rather to the gaps in our knowledge
than to the disconnected remains known or preserved by chance.
It further seems to him that by adopting this point of view,
a man of the North abjures what is in his own blood for something
foreign and acquired. Non-representational art is not more
backward or more primitive than representational ; it is simply
different. Instead of being proud of what we have done, we of
the North ought rather to deplore our excessive surrender to
the histrionic feeling of the South. To personify or anthropo-
morphize all and everything is to attempt the opening of every
door with the one master-key ' Man ', and to recede far indeed
from great Nature and her secrets. It is clear enough that the
present generation has deliberately turned against representation.
The Greeks introduced the human form into their art at a very
early period ; it was their good fortune that they did it without
ulterior motive ; the military states which had gone before them
knew only too well why they gave a common form to God and
King. Christians were once as far removed from representation
as the Greeks were originally. The assumptions which underlie
a fact consistently ignored have now to be considered.
Among historical forms of religion, Monotheism appears
to favour a non-representational system. Judaism and Islam
are examples ; we shall see that in its origins Christianity inclined
in the same direction. If a non-representational method really
meant no more than an elementary and undeveloped phase, it
could never hold its own by the side of representational religions.
Had it been only this, Hellenism would no more have accepted
a non-representational Christianity than it accepted Islam, and
the old representational art of Mesopotamia and Iran would never
have tolerated a non- representational and popular Mazdaism. We
shall come back to this point. The truth is that the sentiment of
nomads and northern peoples resents the entrapping of religious
REPRESENTATIONAL ART 105
impulse by means of representation, which they consequently
exclude from their religion. I conceive the whole North of
Europe and Asia as a vast funnel-shaped area, narrowing to
a point in the region between the Black Sea and the Altai. The
Aryan peoples who made their way southward through this
tract carried with them the northern artistic feeling, which
also traversed the whole South from East to West, from Central
Asia to the west coast of Africa. In my book Altai-Iran I thought
it possible to approach the art of these regions from the side of
Islam. From the Syro-Egyptian angle, once the starting-point
of Christianity, I felt my way towards the North-east, the quarter
from which the whole artistic movement of Islam flowed back
like a tide. If I had to begin again, I should prefer to start
directly from the Northern side ; but as yet such a course might
be premature, and in the present section I shall keep to
the path already trodden. In a later section we shall learn some-
thing of a representational cross-current between Asia Minor
and India.
1 . Islam
Islam has no religious representation. Starting among the
nomads of Arabia, it broke through the Hellenistic-Indian
barrier in a north-easterly direction, and soon ceased to build
with stone, after the fashion of Christianity, which preceded
it in Syria, Asia Minor, and Armenia, adopting instead the
Mesopotamian and Persian material, brick. It covered its brick
walls in an un-Hellenic and un-Indian manner with geometrical
repeating patterns, which originated in the use of diverse materials
and crafts and produced their eff"ect by openwork, by slant-cut
surfaces, by sheen or colour, never by the plastic representation
of natural forms in light and shade. All these patterns, all these
kinds of work came into Western Hither Asia through Persian
influence just as representation had entered the East through that
of earlier Greece ; they in no way resulted from any gradual
transformation of classical feeling, or any change of late-classical
taste in the East. Greek architecture let its forms develop
gradually like the human figure, so that formative art and
construction harmoniously advanced together ; but Islam adorned
the walls of places of worship with surface-filling design, not
admitting the human figure even as a means of giving the scale
for the parts of the building. Despite all religious schisms, the
io6 RELIGIONS WITHOUT
mosque at no time and in no place admitted representational
pictures whether disposed in friezes after the Semitic fashion,
or, as in Pompeii, as centre-pieces in the middle of a wall ;
the only Pompeiian style which we can at all compare with the
later Islamic manner is the architectural, in which the whole wall
was treated as a unit, or the style which counterfeits marble
linings. The Hellenistic painter imitates lining in his frescoes ;
Islam never follows such a course. Rejecting objective form, it
creates patterns that fill borders with rhythmic undulating lines,
and larger surfaces with straight lines forming lattice, net, or
interlacing diaper. The only Islamic ornamental form with
objective meaning is the band of decorative Cufic lettering.
This originated in the old Semitic country, but, like the builder s
art itself, only received farther East the decorative transformation
which gives it its artistic value.
I now treat as briefly as possible the art of the Mohammedan
religious building, or mosque. Down to about A. D. looo, the
period which alone concerns us, the mosque was an open court
surrounded by walls with roofs on the inner sides, the roof
towards Mecca covering more ground than the others. At
Medina, palm-trunks formed the original supports ; in Syria
and Egypt columns from Christian churches were employed,
connected after the Arab manner by arcading, the roof being
formed of any suitable material that came to hand. Under such
conditions as these, there was .small scope for an art with indi-
viduahty, going beyond the simplest practical needs, either in
the home of Islam, or in Syria and Egypt, the first countries into
which the faith of Mohammed spread. If embellishment was
desired, craftsmen were pressed into the service wherever they
could be found, either in the district where the building was being
erected, or from richer places such as Coptic Egypt or even
Constantinople. Corporations of workmen for ordinary purposes
were everywhere at the disposal of the conquerors.
It is characteristic of these times that Islamic art first attained
individual expression in Persia ; in the present place the bare
fact can only be stated, the reasons for it will occupy us later.
As long as the Ummayads in Syria were endeavouring to outbid
the rulers of all countries, monuments were erected which betray
their foreign inspiration in every line. In the Dome of the Rock,
built to supplant the structure round the Kaaba at Mecca, the
rock was enclosed within a circle of columns taken from a church,
REPRESENTATIONAL ART 107
and covered with a wooden dome ; this was in its turn enclosed
by a second circle of columns and an octagonal outer wall. The
original decoration of the interior is represented by the spandrel-
mosaics which reproduce purely Persian designs. In the great
mosque at Damascus a central dome was flanked by aisles ;
here too only the long members were modified in a Christian
sense by the use of columns ; the structural type itself is repre-
sented on Iranian soil by an earlier building, Eiwan-i Kerka,
or Tak Ivan. The dome mosaics show ' landscapes ', com-
binations of trees and other features with only a remote resem-
blance to nature, and intended rather to symbolize than to
represent. Thus even in this early Syrian period these exceptional
mosques, which differ so markedly from the usual Mediterranean
type, were already penetrated by Persian influence.
The earliest group of systematically planned mosques
appeared in Mesopotamia ; we may recall the fact that this was
the country which originated the barrel-vaulted building of
Christian times. Islam retained the open court of its earliest
mosques, but the supports for the surrounding roofs could be
no longer provided by columns collected at haphazard from
Christian churches ; they had to be built up brick by brick.
This brought order into Mohammedan art. Mesopotamia was
the first place where surviving monuments show what we should
call uniformity in decoration ; we see bands of ornament in stucco,
carved or stamped, like the familiar designs in the mosque of
Ahmed Ibn Tulun at Cairo, built in A. D. 872. It has been
persistently maintained that, like the wooden panels carved in
the same style, they are of Egyptian derivation. I fancy the
excavations at Samarra have sufficiently shown, despite the
desperate efforts made even here to prove the contrary, that in
all this incredible wealth of borders and broad surfaces filled with
geometrically planned designs of formal scrolls in lines or
repeated in diapers, the East was the sole source of inspiration.
It has already been suggested that certain materials, like
stone and wood, suffice in themselves to produce a decorative
effect, while others, notably unburned brick, require facing
to make them pleasing to the eye and to secure durability.
Such facing may be carried out in stucco, tiles, mosaic, or other
similar means. People are ready enough to admit that even
mosaic has an * oriental ' character, though in their view the
essential point to notice is that the oriental could make little of
io8 RELIGIONS WITHOUT
his own invention : Greek wits were needed for its development.
But before venturing upon such judgements we ought first to ask
ourselves whether this mode of decoration was not intended rather
to produce a general effect from a distance than to represent par-
ticular things through compositions designed for a nearer view.
If so, should not quite different standards be employed in criti-
cizing two distinct methods ?
The decorative facing of walls was known in the ancient
East, and naturally in the first place in Mesopotamia. But there
representational compositions were employed, running along the
walls in horizontal zones. This is a style which cannot have
originated in wall-lining, but must have been superimposed.
It IS essential to wall-lining that the wall must be regarded as
a unit, that is, a framed surface to be filled with an ornament
befitting this conception of its nature. In their ideas of wall-
decoration North and South stand over against each other like
distinct worlds ; the South bases everything upon representa-
tional art, the North begins with craftsmanship. For the moment
no more need be said of the representational system; we shall
revert to it in Chapter VII. In the present place I confine
myself to ornament, the evolution of which was due partly to
physical conditions, to the treelessness of Iran ; but in a greater
degree it must be ascribed to a religious belief, the creed of
Mazdaism.
II. Iran
In the late period with which we are concerned, we must
recognize fundamental contrasts in art, similar to those which
existed in a remoter antiquity. As we distinguish between
Egyptian and Mesopotamian art, the one characterized by an
organic treatment of architecture, the other by the habit of
covering the walls of buildings with decorated linings, so we
should now discriminate between the organic architecture of
the West-Aryan Greeks, and East-Aryan (Iranian) architecture,
with its lined and decorated surfaces. In the intervening Semitic
tract of Mesopotamia, an uncertainty of procedure may be perhaps
detected ; Greek types may appear, transformed in an Iranian
sense. But except in the Sassanian region in the South, national
feeling in Iran prevented the intrusion of Greek forms on any
extensive scale ; it excluded the columnar style, acanthus
ornament, and, above all, the representational system. After
REPRESENTATIONAL ART 109
Alexander, there was a perceptible infiltration of popular orna-
mental methods from Iran into the Mediterranean area.
The introduction of wall-lining is the most conspicuous
instance. This was probably adopted by ancient Egypt from
Asiatic sources. About 280 B.C. the so-called style of incrusta-
tion appeared in Alexandria, essentially the same as of the First
Style in Pompeii ; this style uses the slab of coloured marble
as the slab of tufa is used in Armenia. It was soon succeeded
by the architectural wall-paintings of the Second to the Fourth
Pompeiian styles, which came westward via Antioch and Rome.
These are fantastic architectural motives giving quantity without
quality after the fashion of much Indian work, and distributed
over the surface almost like repeating patterns. Semitic and East
Aryan art were never associated with these extravagances. Their
wall-lining and ornament remained independent of structural
forms, creating their own laws on the basis of material and crafts-
manship. Their place of origin, as I showed in Altai-Iran, was
the north-east of Iran.
The Persia of the Early Christian period must be distin-
guished from the Persian State of to-day. It included Mesopo-
tamia, which in Sassanian times (a.d. 226-640) had even the
honour of including within its bounds the capital, Seleucia-
Ctesiphon, situated m the territory of the ancient Babylon. We
have therefore to inquire whether this inclusion favoured the
spirit of the ancient East, or whether after all the advantage lay
with the intruding spirit of Iran.
It would appear that at this period Mesopotamia was just
as uncreative as Egypt ; in architecture it had the barrel vault,
in sculpture and painting its representational system. In the
province of ornament, we may infer from the Parthian palace of
Hatra, and from certain Christian churches to be mentioned
below, that Hellenistic forms were modified to suit an art of
surface decoration ; deprived of their organic meaning they were
confined to bands or ' friezes ', running in broken courses in
all directions. There was no creation of new forms. — ?
The plateau of Iran is divided by a salt desert into two
absolutely distinct parts, a northern, the seat of Parthian power
between 247 B.C. and a. d. 226, and a southern, the country of
the Sassanian dynasty (a.d. 226-A. D. 64P). The two parts
were connected on the east and west sides by the Indian border-
lands and by Mesopotamia respectively.
no RELIGIONS WITHOUT
We are familiar with the art of Southern Iran through
important surviving reliefs representing Achaemenian and Sas-
sanian monarchs. In the gigantic barrel-vaulted hall of the
Tak-i Kisra, that Mesopotamian wonder of the world, with its
fa9ade of blind arcades, we recognize the remains of a Sassanian
palace, though nothing survives of the interior decoration. In
neighbouring Persian districts are other ruins of similar style,
though upon a much smaller scale.
The key to the problems involved is provided by early
Mohammedan buildings in Mesopotamia, revealed to us by
modern excavation. Remains of immense structures have been
laid bare at Samarra telling us at last something definite as to
the style and the appearance of the buildings which we
may expect to find in the south of Persia : palaces and
mosques with vast walls and piers of burned brick allow us
to infer a despotic atmosphere in which the passion for
magnificence could hardly be excelled by the Roman or the
Byzantine courts in the utmost display of their arbitrary power.
The excavations at Samarra brought to light a whole city of
palaces and mosques extending by the Tigris over an area
33 km. in length by 2 km. in breadth, all built in a uniform
style during the ninth century, and soon afterwards totally
abandoned.
Here we find a wealth of wall-decoration. For the most
part this consists of repeat-patterns on the stucco lining, either
impressed in slant-cut surfaces by means of wooden stamps, or
deeply undercut by hand to produce sharp contrast of high light
and black shadow. Hardly any of the designs are indigenous
to Mesopotamia ; those with slant cutting come from the Altai,
those with vine-derivatives from Persia. Iran, its northern and not
its southern area, was the real source of this decoration. Islam
replaced the Mazdean religion of the Sassanian state. It is a point
of crucial importance that the fundamental and permanent qualities
of Mohammedan art were derived neither from Damascus nor from
Baghdad, but from the area of Altai-Iran. My book upon the
art of that region was written to establish this point. Here I
need repeat but little from its pages, turning rather to the religious
aspect of the problem, an aspect to which I was already at that
time able to allude.
We have recognized in Edessa and Nisibis one religious
and intellectual centre of Christian Oriental art, from which
REPRESENTATIONAL ART in
a Semitic influence was diffused throughout the world. But
there was another and remoter centre on the frontier dividing
Eastern from Western Asia, the influence of which, passing
Pamir and Altai, extended by a northern route into Buddhist
China. During the first millennium its position as an area where
Indo-Aryan and Turco- Mongolian influences crossed lent it a
significance the beginnings of which I sought to trace back to
prehistoric times. Indo-Aryan elements are still to be detected
in Islamic technique and Islamic designs, derived from wooden
construction, Turco- Mongolian in motives of decoration bor-
rowed from tent-coverings and from personal ornaments of
metal.
Altai-Iran was confined to a study of ornament. My work
Die Baukunst der Armenier und Europa rounded off the problem
from the architectural side. We can only explain the Armenian
building with dome on squinches over a square plan on the
hypothesis of its introduction by the Arsacid royal house and the
partly Oriental Nacharars, who used Trans-Oxanian constructional
forms for their baths, palaces, and tombs. The region across the
Oxus and Iran are the home of the dome on squinches over a square
plan, and there it persists in the domestic architecture of to-day.
In some Iranian villages scores of such domes are to be seen.
Their distribution extends to the Buddhist temples of Chinese
Turkestan ; it includes the palaces of Seistan and Persis, where
the two well-known examples of Firuzabad and Sarvistan
illustrate North Iranian domed construction as developed on
South Persian soil ; a whole complex is held in equilibnum by
the juxtaposition of individual domed chambers and barrel-
vaulted halls. Only in North Iran, at Bus-i-Hor, can we show
to-day the single dome in rubble-concrete with tiled roof, a type
fulfilling all the conditions required for a further development in
Armenia. In the province of decoration the sequence is no less
plain than in that of architecture.
My starting-point is here once more the angle between
Altai and Iran, the meeting-ground of Indo-Aryan and Turco-
Mongolian influences. Let us recall the cardinal fact that both
were non-representational. The Indo- Aryans soon surrendered
to the contrary practice of the South ; in Mohammedan Persia,
Chinese and Indian representation is frequently found. The
Turco-Mongolians, on the other hand, remained, with few
exceptions, true to their old non-representational methods down
112 RELIGIONS WITHOUT
to the penetration of their territory by European influences in
recent times. During the first millennium devotion to non-
representational art was a trait of the genuine Northern nature
among the Persian and Turkish populations ; it was this which
rendered possible the union of the two artistic streams, that of the
Northerners of Iran and that of the nomad herdsmen. From this
source sprang the power of a movement which came into the
light of history from the East, a movement which succeeded in
maintaining itself against that art of Southerners, Indo-Chinese
and Semitic-Hellenistic, both, as products of a hot-house culture,
making representation their principal aim. Once more Armenia
gives us a glimpse of the period before Islam, although this
mountainous volcanic country did not adopt the Iranian methods
of covering blank walls. The Armenians preferred to use the
tufa and lava of their own country in place of stucco, glazed
tiles, metal, or carved wood ; a lining of plain stone was their
sufficient decoration. Yet in certain places on their buildings
we find employed forms of ornament characteristic of East
Iran, vine-scrolls, pomegranates, geometrical scrolls, and inter-
lacings of bands channelled with two or with three slant-cut
grooves, all developed almost to the point at which Islam took
them up to form polygonal designs, or, by constraining the vine-
scroll to geometrical law, the so-called arabesque ; such places
are the flat bands of arcading and sloping surfaces under the
eaves, the arcading round the windows, and, later, the blind
arcades covering the lower walls and the drums beneath the domes.
Other countries of which stone is the natural building material
have, like Armenia, preserved such motives by translating
them into this durable substance. While elsewhere we have
only sporadic examples, Armenia has imprinted a uniform national
style on monuments distributed in comprehensive groups.
Armenia is not alone in furnishing us with data for the
mental reconstruction of East Iranian art ; there exist also
a number of widely scattered stone monuments which impress
us as foreign to their environment, and permit us to infer their
derivation from Iranian buildings of unburned brick lined with
stucco, tiles, mosaic, metal, and wood. Two such buildings
are extant in the country east of the Jordan, and one in India.
Of the first two, one, of quatrefoil plan, is in Moab, in the
citadel of Amman ; » its interior is decorated with three tiers of the
1 This building was unfortunately much damaged by bombs during the war.
REPRESENTATIONAL ART 113
blind arcades which were so common on the exterior of Armenian
churches, and were no doubt originally filled with painted orna-
ment. At Amman they are filled with those formal tree designs
which form one of the richest features of Iranian art. We may
reasonably assume that the building was originally domed ;
but as the dome was supported on squinches it has collapsed
without leaving a trace above the four walls from which it rose.
The second of these buildings in Moab is Mshatta. The orna-
ment of the entrance wall at the west end of this three-aisled trefoil-
ended building, four piers supporting three arches linked by
a r-i-shaped moulding above them, and the decoration of the
triumphal arch before the trefoil itself, are so many proofs of
Persian influence. The Corinthian capitals of the entrance wall
are of course Greek ; but the continuous moulding framing the arch
is Iranian, as are the six enclosed rosettes, the four ribs on the
soflit of the triumphal arch and the impost capitals encased in
vine-scrolls. On the other hand the arch's vine-leaf moulding
surmounted by acanthus belongs to that mixture of Iranian and
Graeco-Mesopotamian art which is the distinguishing feature
of the colossal fa9ade now preserved in Berlin.
The chronology of Mshatta would not be a matter of dispute
if archaeologists and art students would only give full and careful
consideration to my book on the subject. At the present time
a comparison with the chiirches of Mesopotamia and Armenia
dating from the fourth to the seventh centuries, and with the
remains excavated at Samarra of the ninth century (all of which
were still unknown at the time I wrote my Mshatta in 1904), enables
us to conclude with certainty that Mshatta antedates all other
Christian and Islamic monuments, and may even be regarded
as perhaps of Parthian rather than Sassanian origin. The classical
friezes (comprising base mouldings, cornice, and an intermediate
band of zigzag) should be compared with those of Mesopotamia,
where the oldest surviving example belongs to the year a. d. 359 ;
the rosettes in the zigzag should also be compared with Armenian
parallels ; and the vine-scroll with enclosed animals not only
with that which occurs on the so-called throne of Maximianus,
but also with classical examples on the one hand and the pilasters
of Acre and Zwarthnotz (a. d. 650) on the other. Such compari-
sons will clearly show that the Mshatta fa9ade cannot be
attributed to the early Islamic period. Indeed it reveals that
fusion of Iranian and Greek art which succeeded the displacement
245' I
114 RELIGIONS WITHOUT
of the latter in late Roman times, and led gradually to the develop-
ment of Byzantine art on the Mediterranean, of * Romanesque '
in the West, and to the complete triumph of Iranian art in the
world of Islam.
The third stone structure with Iranian decoration is the
Sarnath Stupa near Benares (Fig. 32, opposite). It has borders
of scroll-work, in which the vine is replaced by the lotus, while
the intermediate zone is occupied by the same kind of swastika
fret which occurs as a continuous pattern on the columns of the
Amida fafade and on the remains of numerous Christian churches
in Egypt. Sarnath and Aswan (Assuan), both of them situated
near the tropic, were probably the two most southerly points
reached by Northern or mediaeval art.
Examples of Iranian decoration in wood and stucco from
the eighth to the tenth centuries have gradually come to light
in great profusion. A comparison of the stuccoes of the Church
of El-Hadra in Deir es-Suryani in lower Egypt (inspired by
North Mesopotamian art) with almost contemporaneous work
in Afghanistan, which takes us back to Mahmud of Ghazna,
clearly indicates the centre from which both were derived, and
even suggests that the influence of Eastern Iran penetrated as
far as Eastern Europe and Scandinavia along the paths followed
by thousands of Samanid coins. The ornament of the wooden
carts and sledges discovered in the Oseberg ship in Norway
shows close affinities of style with East Iranian decoration.
The Samarra excavations and their bearing upon Islam
have given us some idea of the light that excavations in Iran
might reasonably be expected to throw on the relations between
Mazdaism and Christianity.
Glazed wall tiles of Iranian style, dating from about the
ninth century of our era, have been discovered at the monastery
of Patleina near Preslav in Bulgaria. Their ornamentation
presents a close parallel to that of certain capitals found in Iran
(Figs. 45-47, facing p. 146) and of the silver pouch-shaped plaques
with repousse design excavated in Hungary, believed to date back
to the original occupation of the country by the Hungarians.
Persian stucco- workers, like those of Italy at a later date,
would appear to have traversed the whole of the late classical and
Early Christian world. In Cividale I succeeded in finding examples
in the style of Mshatta. In other materials, too, such as textiles
and leatherwork, we find a constant recurrence of Iranian
32 Sarnath, stupa ; zone of ornament ; photo., E. La Roche.
See pp. 114, 125.
33 Island of Achthamar, Lake Van, cruciform church ; South view.
See p. -1 59.
c
J3
-a
a
•a
REPRESENTATIONAL ART 115
features throughout the Eurasian continent. To mention only
a single instance, it is instructive to compare the ornamental
capitals, illustrated in my book on the miniature-paintings of
' Lesser Armenia ', with the decorative art of the Amur region ;
the common source from which both derive must be sought in
the area between the two countries in Altai-Iran.
III. Mazdaism
It has already been pointed out that our total ignorance on
the subject of Mazdean art is not a sufficient reason for denying
the possibility of its existence. Equally inconclusive is the fact
that the cult of Mithra, in its triumphant passage through the
Roman world, borrowed Greek forms in the representation of the
god at the bull sacrifice. The disposition of the, Mithraic temple
and certain traces of its decoration should alone impose caution ;
one may fairly assume that the reason why Mithraism borrowed
from Greek art was that it had never itself depicted the figure
of Mithras, and was indeed generally unacquainted with repre-
sentational art. But that is no ground for proceeding to argue
further that Mazdaism had no art. The apparent absence of
concrete evidence is no reason for doing so.
On this analogy, if our knowledge of Christian art were
limited to the Roman catacomb paintings — ^which belong to
classical Christian art — we should be forced to conclude that the
Christians had no distinctive art of their own, simply because
they lacked a distinctive style of representation. In point
of fact, the only pre-Constantinian examples of their art at
present known to us are those at Rome, which are Alexandrian.
But it is a fallacy to conclude, as is so often done, that other
Christian monuments cannot have existed before the time of
Constantine, simply because they have not survived to the
present day. In the study of Christian architecture attention
has hitherto, for some inexplicable reason, been confined to
the Hellenistic timber-roofed basilica, which is structurally just
an example of classical Christian art. The preceding chapters
of this book will, I believe, brine about a clearer perception of
the facts. There must assuredly have been some form of
Christian art in the East during the first three centuries of our
era, an art of Christian communities and their places of assembly.
And similarly we may fairly assume that in Achaemenian, Parthian,
I 2
ii6 RELIGIONS WITHOUT
late Hellenistic, and Sassanian times there existed some form of
Mazdean popular art, the character of which we could certainly
never deduce from a study of the grandiose art of the Sassanian
court.
The question still remains whether this new province, into
which I have opened up a path by the method of retrospective
inference from the art of Islam, should be called * Mazdean ',
or whether we should not do better to retain the geographical
term * Altai-Iran *. Non-representational art is characteristic
of the pastoral-nomadic and Northern peoples. It is present
both in Islam and in Christian Armenia. Is it likely that Maz-
daism, the oldest of the religions of Hither Asia established by
a Founder, should not have preceded them on the same path ?
Mazdaism, while exceptionally rich in religious ideas, never
strove in its popular art after realistic self-expression by means
of the human figure. In Christian art we find representational
subjects based upon Mazdean ideas ; but in Mazdean art itself
we shall look in vain for scenes from the life of Zoroaster analo-
gous to the figures of Christ or Buddha in Christian or Buddhist
art. Nor need we expect to find Mazdaism dramatically expressing
by means of the human figure those general ideas of the world
and life, of death and a future state, in which the A vesta is so
exceptionally rich.
It would therefore be wrong to place Mazdean art on the
same footing as that of Buddhism or of Christianity. Nevertheless
research should be able to define this art with some approach to
accuracy. It is impossible as yet to distinguish in particular
cases between religious and profane monuments. In this respect
Mshatta, Amman, and Bus-i-Hor present unsolved riddles. In
the case of Islam, where proximity gives a clearer view, we know
that religious art, though forming an important section of Islamic
art as a whole, is nevertheless far from co-extensive with it.
It is best to apply the term ' Mazdean ' to Iranian sacred art,
because it comprises the religious expression common to Achae-
menians, Arsacids, and Sassanians, whereas the term ' Persian '
suggests a Southern predominance non-existent in the case of
formative art.
Concrete examples of Mazdean sacred buildings are entirely
wanting. I do not propose to consider here whether this fact
indicates that they were always few in number, or whether it
simply results from the perishable nature of the unburned bricks
35 Kairwan, wooden mimbar ; pierced interlacing, niches, &c.
See pp. 117, 125,241.
REPRESENTATIONAL ART 117
of which they were constructed. One fire-temple has been
excavated at Susa, and there is literary evidence for others of
a later date. It is at least noteworthy that the deficiency is one
which Mazdaism shares with the older Oriental cultures of
Mesopotamia and of Iran ; in the latter case, evidence of temples
has only been revealed by recent excavations carried to a great
depth. Similar results will perhaps be achieved for Mazdaism
in the future. At present our knowledge is in the main limited
to its fire altars. And it is a fact of no small importance that the
two oldest examples, the great altars of the rock-reliefs at Naksh-i-
Rustam, already show the blind arcade, that commonest of
motives in the decoration of Armenian churches (Fig. 2, facing
p. 62), a feature which some authorities would derive from Iran,
or from the galleries round early Aryan wooden buildings
(Figs. 15-16, facing p. 80). Before considering this point more
closely I may recall another trace of Eastern influence. An
examination of the evidence clearly proves that orientation in
Christian churches was indigenous to Armenia and Asia Minor,
and spread from these regions, superseding the ancient classical
practice, which made the important end of the building face the
west. One cannot help wondering how far Mazdean influence
may not have contributed to this result.
Mazdean decoration is purely formal ; that is, it entirely
avoids expression by means of human figures. Instead of these
it employs animal and bird forms, scenes from the chase, and
the conventional landscapes common in mediaeval art. But
it has a special preference for friezes of arcading, blind arches,
interlaced bands — in fact just those features which are character-
istic of Islamic art (Fig. 35, opposite) and of that Western
art which we misname ' Romanesque '. It would thus seem
incumbent upon students of art to give more attention to this
forgotten province, if only for its bearing upon the decorative
forms affected in mediaeval Europe.
Of the features properly belonging to Mazdean decorative
art, only a portion can be traced to genuine Northern tradition.
The blind arch, preferably of horseshoe shape, the cushion
base, and its colonnette might be shown to derive from wooden
architecture ; the same applies to the decoration of the walls
beneath the eaves with arcading in low relief or the inclined
surfaces with interlacings. I have treated this question in
greater detail in my work on Armenia. At present I am more
ii8 RELIGIONS WITHOUT
particularly concerned with the patterns used on the inner surfaces
of vaults. They naturally owe their origin to those countries
where bricks were used, namely, Iran proper, Irak, and Mesopo-
tamia. I shall now analyse the designs employed, but only from
the point of view of their religious content.
A. The Hvarenah Landscape.
At the very heart of Aryan piety on Iranian soil lay the idea
of Hvarenah. Soderblom has shown in his work. Das Werden
des Gottesglaubens, that it represents the crowning glory of Iran.
A comprehensive description of its nature is given in a long
hymn in the Avesta. It is connected with the cult of the dead,
representing the might and majesty of departed spirits. Hvarenah
is the power that makes running waters gush from springs,
plants sprout from the soil, winds blow the clouds, and men come
to birth. It governs the courses of sun, moon, and stars.
Hvarenah therefore is that which permeates the whole countryside,
and particularly the land of Seistan through which the river
Helmand flows. To this country, alternately frozen and parched
by drought, water forms the very pulse of life and is full of
a mysterious beneficent potency. ' O water, to him who sacrifices
to thee vouchsafe thy glory ! ' Hvarenah, created by Mazda, is
the power that makes the waters rise from the world-ocean.
From Hvarenah the sun derives his strength. If ' the mighty
immortal Sun with swift steeds illuminates and warms ', if it
purifies earth and water and banishes the evil demons of darkness,
it is from Hvarenah that such magic influence comes. Even so are
Moon and Stars endowed with power and fulfilled with majesty.
Were such ideas expressed in art, we should expect to see
a barren landscape, above it the sun with his swift steeds, below
it the world-ocean ; the land between would have its gushing
springs, and its scattered plants springing from the earth ; over
all would float clouds. ' With milk (cattle) have I created the
'glory of the Aryan people, rich in herds, rich in lands, rich
* in glory ; endowed with wisdom, endowed with possessions ;
* bringing gluttony to nought, bringing hostility (enemies) to
' nought.' In such a Mazdean landscape I should expect to
see flocks and herds and symbolical presentation of attacks
and conflicts of wild beasts.
Are there such landscapes ? If so, they would not be
REPRESENTATIONAL ART 119
representations in the strict sense, but compositions pieced
together out of the elements enumerated above, mere symbols
of nature, devoid of realism. Other artistic treatment could
hardly be looked for in Iran. I thus find myself on the track of
a kind of Asiatic landscape, originating, like the Chinese ' philo-
sophical landscape ' of the Sung period, in religion, or rather in
a philosophy of the universe, and based upon significance and
form, not, like the landscapes of the Southern peoples, upon
natural objects exactly reproduced. It is quite possible that
this type of landscape first passed from symbolism to true
representation not in Iran, but in the regions of Hellenistic or
Christian culture ; in the same way, Indian ideas appear to have
found their earliest representational expression in the Taoism
of China and the indigenous landscape paintings which it inspired.
We shall see at a later stage whether the existence of such land-
scapes can be proved in Christian art. I shall for the present
exclude all mention of Hellenistic landscapes, although they
offer a promising field for research.
Scenes from the Chase. The chase was one of the pleasures
of the Mazdean paradise and future world. We should, therefore,
expect to find it frequently depicted in Iranian art, and not
exclusively in the traditional Assyro-Babylonian style. The
palaces of Persian kings were once filled with subjects of this kind.
A single example of this Court art has been preserved, a rock
relief at Tak-i Bostan, representing the chase of deer and of wild
boars. It consists of flat composition aiming at strict naturalism
rather than at artistic values such as spatial unity ; in this
quality, as well as in its concentration on the person of the ruler,
it resembles the ancient oriental hunting-scene more closely than
the Greek. On Sassanian silver dishes we find similar scenes
repeated in almost endless variety, depicting the chase of deer,
ibex, and wild boar, with the aid of falcons or trained panthers.
The decoration of the main hall in the small castle of Quseir
'Arnra is executed entirely in the Iranian spirit, and shows the
persistence of this ancient tradition in Ummayad times ; the
walls are covered with scenes from the chase. Even the Normans
still decorated their interiors in Palermo with hunting-scenes of
this type. One of these, executed in mosaic, is preserved in
the Palazzo Reale. Later examples are to be seen in the
Alhambra. On the far side of Iran, in Turkestan, this kind
of landscape encountered Chinese influence, and the product of
120 RELIGIONS WITHOUT
their combination is seen in lozenge-diapers with characteristic
conical mountains, sometimes containing animal figures, some-
times real hunting-scenes, or figures of Buddha. The Shosoin at
Nara (a. d. 749-56) contains excellent examples of Iranian work
in both the freer and the more conventional style.
River Landscapes. In the paintings of Chinese Turkestan
the river landscape is a stock composition introduced on floors,
walls, and ceilings ; numerous examples will be found in the
pages of Griinwedel, Le Coq, and M. A. Stein, where illustrations
are given of the cave of the Hippocamps at Ming-oi near Kyzyl,
a fresco in Dandan-Uiliq, and in particular the roof of the
Naga-cave at Sorcuq (Fig. 42, facing p. 137). Here again we can
find parallels on the opposite side of Iran : river-landscapes with
formal tree designs (candelabra motives) similar to those at Sorcuq
occur in the church of S. Costanza and in two Roman apse-
mosaics, of all of which I shall have more to say presently.
These river-scenes, probably Iranian in origin, afterwards
formed a permanent foreground to Buddhist representations of
' The Western Paradise '. The Padmapani tree of Bazaklik,
which I shall presently discuss, is also depicted as growing out of
waves (Fig. 37, opposite).
B. Hvarenah Symbols.
Mazdaism, like Buddhism, and Christianity at a later date,
appears to have made full use in its symbolical art of that great
stream of animal ornament which began in the North and East,
poured in flood through the funnel-shaped gap between Altai and
Iran, and finally issued in Hither Asia. The clue to the interpreta-
tion seems here again to be Hvarenah, the Majesty of God,
the best possession of the home, averting evil, giving increase
to prosperity. Animal figures are repeatedly named as vehicles
of Hvarenah. ' In mythology and popular belief Hvarenah
* appears sometimes as a bird in flight, sometimes as a swimming
' or diving creature, sometmies in other animal forms, and
* follows the Elect (after the Karnamak and the Shanama the
' glory of the king came to Ardeshir in the likeness of a great
* ram and rode at his side). Sometimes it resides in the sedge
' at the lake-side, is eaten by the cows and passes into their milk.
' The high esteem in which the cow was held in Iran, as in
' India, rendered it an appropriate medium for the transmission
37 Bazaklik, Chinese Turk-
estan, temple group ; painting ;
Padmapani on scrolled stem ;
after A. Griinwedel. See pp.
120, 123.
36 Brunswick Museum, ivory
carving; Yima. See p. 122.
38 Spoleto, fa9ade of S. Pietro ; photo., Alinari. See p. 121.
REPRESENTATIONAL ART i2i
* of divinity.' The bird, the ram, the sedge in the lake, the cow ;
are these really unfamiHar to us in art ? I think we have here the
key to many riddles which have perplexed our minds but have
always been left unsolved.
Let us examine a facade such as that of the Church of S.
Pietro at Spoleto, a place full of interest for students of Early
Christian and mediaeval art (Fig. 38, opposite). All periods,
including recent centuries, have left their mark on the building.
But there still remains a central core, represented above all by
sculptured ornament which one critic ascribes to masters of the
year c. 592, others to the Umbrian masons of the period of
transition to Romanesque. Both the main entrance with its
horseshoe arch and the pair of cows in the lower panels of the
pediment impress us alike as foreign to their actual environment.
Noteworthy, too, are the decoration of the blind arcades on either
side of the main door, the animal scenes in the adjacent panels,
and several other features into the significance of which I cannot
at present enter. Are not these things echoes of a forgotten music
audible here upon Italian soil ? I shall now consider the various
Mazdean symbols individually.
Animals. A well-known ornament of the Sassanian royal
crown is the pair of wings. Plastic art adopted them as an
ornamental feature, for instance, in the entrance fa9ade of
Mshatta. That these may have been actual wings attached to
the crown is suggested by the passage in which we read that the
feathers of the bird Varegan are protective against the black
magic of an enemy. No man carrying a single one of this bird's
bones or feathers can be either killed or overcome : * He first
receives homage, first majesty.' Probably the disk between the
wings of the Sassanian crown originally represented the sun ;
it is sometimes placed above a crescent moon. The bird Varegan
is the vehicle of Hvarenah. When Yima in the presence of the
mighty of the empire ascribed to himself the creation and the
happiness of the world, ' the glory flew forth in the likeness of
a bird ' ; the bird Varegan forsook Yima in its three capacities
successively as the glory of the priest, of the king, and of the
peasant, as it is written in the Avesta :
Yima, the radiant, the good shepherd.
Wandered haplessly to and fro ;
He hid from before his enemies.
And concealed himself under the earth.
122 RELIGIONS WITHOUT
In many collections there are ivory carvings representing a good
shepherd who sits dreaming on a hill. Among the rocks are
depicted all manner of creatures on either side of a central spring
or waterfall. A particularly fine old specimen in Brunswick
(Fig. 36, facing p. 120), and the rich example at Naples, depict
in addition the cavern below, in which Yima is seen reclining
before a book doing penance like the Magdalen. Sometimes
sheep are the only animals represented, but frequently we find
various other species both wild and domesticated, as in a marble
relief at Athens. Sometimes, too, they symbolize the triumph
over evil, like the figure of the lion on the hare at Naples. It
is by no means improbable that the Christian conception of
the religious founder as the good shepherd fell upon peculiarly
favourable soil in Iran.
Figures of single animals are distributed so widely and in such
profusion alike in the art of Islam, on the Mshatta fa9ade, at
Amra, Amida, and as far afield as Korea (Fig. 31, facing p. 99),
that a separate volume would be needed to do justice to the sub-
ject. Probably these figures were originally symbols of Hvarenah
and of other Mazdean ideas ; in later times their purpose may
often have been purely decorative. It is remarkable that these
animals and birds are frequently depicted with a branch beside or
underneath them.
Plants. The vine and the pomegranate are certainly to be
included among the symbols of Hvarenah. Such at least is the
inference I draw a posteriori from Armenian churches like that of
Zwarthnotz (erected in a. d. 650), and its imitations. Vine and
pomegranate are here used in combination to decorate the
spandrels of the blind arcades on the exterior of the building ;
in other churches both in Armenia and in Syria they are similarly
treated but not together. I have pointed out in my Mschatta
how frequent and striking a feature the vine forms in Islamic
art ; coincidences between it and the art of Early Christianity
in Hither Asia are probably due to their common debt to Iran,
and I conclude that both vine and pomegranate were Mazdean
symbols. These types are characteristically free from the
familiar realism of Roman work, and it is not surprising to find
the vine-leaf treated conventionally both at Mshatta and earlier
at Hatra, the stem being placed on the top of the leaf and made to
terminate in one or more buds. The branching of the vine is
likewise arbitrarily planned with the sole object of filling a flat
39 Island of Achthamar, Lake Van, cruciform church, A.D. 915-21 ;
East side. See pp. 123, 150, 159.
40 Lemberg, Armenian gospel of a.d. 1198 ; Eusebian canons.
See pp. 123, 241.
REPRESENTATIONAL ART 123
space ; the plant does not conform to the laws of natural growth.
The hunting frieze of Achthamar, which is prima facie strangely
inappropriate to a church, appears to be a pure echo of Mazdaism.
It is particularly striking that the wood, in which the chase is
depicted, is composed of vines and pomegranate stems or
branches (Fig. 39, facing p. 122). I shall not recur to the
* arabesque ' (Fig. 31, facing p. 99) at this point.
The commonest motive is a kind of * tree ', not less widely
removed from nature than the landscapes, but like them composed
of a number of elements, each of which, taken separately, is
realistic. I call this a * formal tree ' or Candelabrum. It occurs
as a border on the outer rock-sculptures of Tak-i Bostan ; from
the thick stem and branches feathery leaves of Indian style grow
upwards like palmettes. The branches terminate in buds
curving to one side, such as are often depicted as pendants to the
horse trappings of riders in early Persian representations. The
distribution of this ' formal tree ' design throughout the Asiatic
countries both East and West of Iran is so uniform as to prove its
popularity in Iran proper, the centre of its distribution. At
Bazaklik in Turkestan a formal tree of this kind forms a support
for a six-armed figure of Padmapani (Fig. 37, facing p. 120);
it might equally well be used as a support for a cross in one of
the countless miniatures of an Armenian manuscript (Fig. 40,
opposite). In Mshatta it appears in the midst of the scrollwork
filling the great zigzag, while at Amman it is used in great variety
to fill the blind arcades of the interior (Fig. 29, facing p. 99).
It is an equally familiar feature of the interior decorations in
Armenian and Georgian buildings, as well as of the south sides
of Georgian churches. Even in modern works of industrial art,
such as an Armenian embroidery from the Bukovina (Fig. 41,
facing p. 124), it reappears in opulent growth on either side of
a hunting-scene interspersed with flowers and animals, be-
traying unequivocal, if indirect, connexion with those formal
tree-designs which occur at intervals along the walls of the nave
in the Church of the Nativity at Bethlehem. Whether this design
is to be identified with the ancient motive of the tree of life and
the Mazdean Haoma is a question only to be decided by scholars
better versed than myself in the archaeology of Mesopotamia
and Iran.
124 RELIGIONS WITHOUT
C. Decoration oj the Mazdean House.
' May holiness, strength and prosperity, excellence and
* happiness dwell within this house.' ' May the Excellence
' that brings fortune never fail within this house.' ^ Such passages
show that Hvarenah was a virtue which men desired for the
houses in which they lived. In the blessing invoked upon the
hearth and home of the pious, the spirits were invited to enter
with their manifold gifts, ' in order to further the reign of might
and excellence.' When, remembering all this, I contemplate
the rich decoration of monuments like Mshatta and Amman,
I begin to perceive that this decoration does not proceed from
a mere delight in ornament, but that it has a symbolical founda-
tion. Fragments of stucco ornament from a building in Mesopo-
tamia recently brought to Berlin have, amid other motives,
medallions with pearled borders (the pearling perhaps repre-
senting pomegranate stones, and enclosing for central subjects
either the pair of wings with an inscription, or the figure of a
ram or an ibex ; there are also formal trees with birds and
pomegranate (?) fruits. This seems to indicate that the interiors
of houses were decorated with symbols of Hvarenah. I am
inclined to apply the same interpretation to the Mshatta facade.
The illustration (Fig. 34, facing p. 115) shows the walls on either
side of the entrance covered with symbolical designs, the vine
interspersed with birds and animals, especially with those
flanking vases as symbols of blessing and protection. Similarly
enriched with ornament, though belonging to a later period, is
the apse-buttressed square building of Amman ; the inside walls
have three tiers of blind arcades filled with Hvarenah symbols
(Fig. 29, facing p. 99). The Korean tomb (Fig. 31, facing
p. 99) should perhaps also be included in this category.
In this connexion I may mention another kmd of interior
decoration, familiar to students by its occurrence in Christian
churches, although its origin has never yet been satisfactorily
explained. I refer to decorative hangings without representational
designs. I conjecture that it was largely by way of Iran that
they made their fresh entry into the West. The influx of pastoral
nomads, especially Turks from Inner Asia, makes intelligible
the spread of influence from that quarter. Chinese Turkestan
is permeated with it. Undoubtedly the tent, which served
* These are the terms used in the Zend Avesta.
04
Oi
o
u
'>
O
1)
•a
'o
u
X>
6
c
2
'c
e
e
3
U
en
3
"o
b
c
3
C
C
1>
REPRESENTATIONAL ART 125
these pastoral nomads for a dwelling, was important to the
art of Iran during the Mazdean period long before the rise
of Islam. I attribute to it the origin of the lambrequin
pattern, perhaps also the use of folded hangings to decorate the
lower parts of walls, a practice which spread rapidly, and of
which we have evidence in the case of Earlv Christian churches.
Painted imitations of it are legion. It appears in the mosaics
crowning the dome in the Baptisteries at Naples and Ravenna,
and is a constantly recurring feature in the roof-paintings of
caves in Chinese Turkestan. We are not yet in a position to
say whether it had any connexion with Hvarenah.
It is possible that the art of Iranian wall-linings may ulti-
mately be connected with the designs of such coverings, varying
with the materials of which these consisted. It undoubtedly
derives its distinctive character from the qualities of woven and
plaited fabrics, from wood and metal ornament with ' slant-cut '
ornament, stucco openwork, &c. Among the designs, that of the
swastika was a particular favourite, and reappears constantly
in a variety of combinations. The finest of all examples of this
kind of work are found on the wooden pulpit of the great mosque
at Kairwan which originally came from Baghdad ; it reveals
an astounding variety of patterns in interlacing openwork
(Fig. 35, facing p. 117). In stone we have the older parts of the
Amida fa9ade, sculpture in Coptic churches, and on the Buddhist
stupa of Sarnath (Fig, 32, facing p. 114). On Iranian soil there
are similar designs on architectural and other remains from the
time of Mahmud of Ghazna ; we found them on the earliest
Islamic remains discovered by us in Khorassan.
D. Mazdean Costume.
We have seen that houses were decorated with symbols
to secure the protection of the divine might and majesty ; cos-
tumes appear to have been similarly treated with a view to
protecting the wearer. In the rock-reliefs of Tak-i Bostan one
cannot help noticing that the garments are completely covered
with symbols of Hvarenah. The commonest of these is the duck ;
and next to it the cock and the curious Persian dragon, viz. a
winged lion with a peacock's (?) tail ; with these symbols occur
the formal tree and the rosette. Remains of silk fabrics have
been preserved, which are true counterparts of these imitations
126 RELIGIONS WITHOUT
in stone. The same designs recur in paintings in Chinese
Turkestan. Moreover, a large number of stuffs have scenes from
the chase resembling those on the well-known silver dishes.
We are thus once more confronted with the art which we have
already defined as Mazdean. Its later influence in Christian
times, which will be discussed later, confirms the view that we
have to deal here not with meaningless ornament, but with
a form of decoration which originally possessed religious signifi-
cance. I refer the reader once more to the seventeenth-century
silk embroidery (Fig. 41, facing p. 124), acquired by the Austrian
Folk Museum in Vienna from an Armenian private owner in
Czernowitz. On either side is seen a formal tree with palmettes
and birds ; in the centre, under a house with two female figures ^
is a hunting-scene disposed in two rows to be followed from left
to right ; it consists of mounted figures holding falcons in their
hands ; beneath them is a gryphon confronted by wild animals
in the midst of which are seen a single figure with a falcon and
a pair of others with guns. The latter give a contemporary note ;
but the remaining portion of the design probably belongs in all
essentials to an ancient I rano- Armenian folk-tradition.
IV. Manichaean Art
Manichaean illuminated manuscripts must have been the
vehicle of a very rich form of decorative art. Mani, who suffered
martyrdom at Gundeshapur in a. d. 274, was himself a painter,
and therefore the only religious Founder from the outset aware
of the importance of the formative acts as a means of influence.
His teaching, comparable with that of Islam in its rapid expansion
and its proselytizing power, soon made its way into the West
also. The presence of this Babylonian religion can be attested
in Rome and Gaul in the fourth century, and it continued to
exert an influence for centuries after. The Albigensian move-
ment in France was indirectly connected with it. Now this
Manichaeanism, which was a forerunner of the Persian movement
in Christian art, must itself have been a carrier of oriental art
forms to the West. Augustine after his conversion bitterly
demanded the burning of the numerous costly manuscripts,
of which he specially notes the fine bindings, but he retains
1 Compare the architecture of the Chinese temples in the reUefs illustrated
by Chavannes in his Mission archeologique dans la Chine Septentrionale,
REPRESENTATIONAL ART 127
the words : non in aliqua mole corporea itispicanda est pulchritudo.
Discoveries in Turkestan confirm this tendency to rich orna-
mentation, although here figure art predominates. The Mani-
chaeans were also familiar with a kind of Hetoimasia (Preparation
of the Throne) in the vacant decorated chair which at the feast
in memory of their Founder they set up as his symbol.
It is probable that Iran was the source of those systems
of decorative design occurring in early mediaeval manuscripts
which cannot be referred to Hellenistic art. How far Hvarenah
motives underlay these designs is at present a doubtful question.
It is none the less a striking fact that in the art of illumination,
which flowered so richly in Armenia and passed thence to the
West and Byzantium, a prominent place is occupied by familiar
Hvarenah motives, such as figures of birds sipping water and
symbols of the divine majesty. The familiar fish-bird initials
of Merovingian and Armenian manuscripts may also derive from
the same source.
These indications suggest that non-representational art won
for itself a serious religious basis in Iran. There is a kind of senti-
ment to which representational art is repugnant on account of its
histrionic character. Let us call this feeling popular, and oppose
it to that of the cultured upper classes, and we shall at once see
what I mean when I contrast the harsh conditions of the North
with the forcing-house atmosphere of the South. Students of
Indogermanic art must start with certain well-defined presup-
positions in so far as the North- and East-Aryans form the subject
of their research, and Buddhism and Mazdaism, Christianity
and Islam, stand out against a background of Asiatic migrations.
The ancient Aryan migrations are connected with a form
of wooden architecture, which was carried by the Greeks to the
Mediterranean and by Iranians and Indians to the Far East,
and had a decisive influence upon the development of architecture
both in these regions and in the North. It was accompanied
by a decorative art originally devoid of figure subjects, but soon
superseded in Greece and India by the representational style
indigenous to those coimtries. It was otherwise in Iran. There
the non-representational style became permanent through the
influence of the national religion, and subsequently passed over
to Islam without substantial change. We shall now have to
consider whether things were not very much the same with
Christianity.
128 RELIGIONS WITHOUT
Be this as it may, it shows a certain bias to derive the origin
of Christian art exclusively from antiquity or so-called classical
sources. This method confines attention to the forcing-houses
of Southern culture, and ignores the pastoral nomads, with whom
were linked those two small nations which gave birth respectively
to Christ and Mohammed, the last two individual Founders of
religions. It likewise excludes those branches of the Northern
races which penetrated to the South without succumbing to
classical culture, and interchanged ideas with the small nations
inhabiting Palestine and Arabia. The religious art of both Jew
and Moslem is steadily opposed to representation. The former
only begins to waver after coming into closer contact with
Hellenism and being thoroughly permeated with its influence,
as for example in Alexandria. The artistic tendency of Islam
was not due in the first instance to the prohibition of figure art
by the Koran ; the really decisive factor was that the Beduins,
like the Jews, were nomads at the time when their religion was
founded. Moreover we know that the religion of Monammed
came with a rush across the desert to Iran, and only found its true
intellectual foundation after contact with the pastoral peoples
of the Altai-Iran region. When we seek the origin of Christian
representational art we encounter the Hellenism of the Mediter-
ranean ; similarly on our present path of research we are con-
fronted bv Mazdaism as the dominant religion. The appearance
of Mazdaism in Armenia allows us to test the conclusions reached
from the above evidence ; we can infer the Mazdean attitude
towards representation from what occurred in that country
during the Christian period. The Sassanian dynasty was hated
in Armenia, and the Arsacids remained in power until the year
428. The Armenian Church, in so far as it remained national
in character, eschewed representation ; may we not infer that
Mazdaism originally did likewise ?
Clearly, Mohammedan art does not stand alone. There
rises before us the possibility that Mazdaism and the earliest
Christian art were pioneers of Islam, and closely akin to it in
spirit : that we have in fact a whole group of non-representational
religions forming an enclave between Buddhist and classical-
Christian representational art. It is perhaps only on this hypo-
thesis that we can rightly understand the beginnings of Christian
Church art, the violence of the iconoclasts at a later time, and
certain characteristic features of Christian art in the West.
REPRESENTATIONAL ART 129
It has long been recognized that Early Christian mosaics are
throughout arranged according to a definite plan, the stamp of
which is most definitely impressed upon the apse. In it the hand
of the Almighty holding the wreath takes the place of the Hellen-
istic radiating bands of colour ; beneath is seen a conventional
* landscape ', in its powerful but rigid expression, equally far
removed from Hellenistic ideas. At the bottom is a strip of
green foreground decked with flowers, behind it other bands
of colour ; are these really intended merely to give depth and
perspective, as Sybel has maintained } The figures of aquatic
animals and birds which fill out the spaces are likewise ascribed
to Hellenistic picture making. But may not the truth be that
the putti form the only Hellenistic contribution to what is essen-
tially a river Helmand, a feature of Hvarenah landscape, in
Egypt called the Nile, and in later times christened the Jordan.
The palms (' which Italian artists could see growing in their own
country *) and the phoenix are symbolic of Paradise. What is
this but a picture of the earth dominated by the might and
majesty of God, whether in the guise of the sun, symbol of
Ahuramazda, or in that of the cross, symbolic of Christ ? There
were apses, such as that described by Paulinus of Nola, in which
the sole decoration consisted of such composite symbolical
landscapes. Can this be properly described as Hellenistic, as
Christian classical art ?
US'
VI
Non-representational Church Art, and
the subsequent Anti-representational
Movement
THE constant features in the religious art of the North
and China during their earliest period of development, as
well as in that of Mazdaism, Judaism, and Islam, appear
to have formed the first stage of Christian art also in the inland
regions East of the Mediterranean. From these regions there
sprang, in despite of Hellenism, a form of non-representational art
extending as far as Rome, Lower Italy, the two Gauls and the
British Isles. This fact has hitherto escaped observation because
the Catacombs with their paintings have been too exclusively
regarded as the only starting-point. There are writers who assure
us that they would gladly trace the course of Christian art from the
East Westwards, but that the complete absence of earlier material
in the East unhappily precludes any such attempt in the case of
the Catacomb paintings. But if this be so we are faced with the
question whether our whole attitude is not based upon a false
assumption and mistaken premisses. Is it really necessary to
place Hellenistic sepulchral art at the beginning of the series ?
Should we not rather concede this place to the earliest forms of
church architecture and its decoration, so far as these may be
established by retrospective inference ? How is it possible to
believe that painting came first, to be followed only at a later
stage by sculpture, and finally by the earliest Christian architec-
tural monuments ? Is it because the view as seen from Rome
wears this aspect, and because in Rome monuments happen to
be preserved only in this chronological sequence ? The classical
archaeologist who proceeds in this manner makes himself entirely
dependent upon the few monuments which happen to have been
preserved. Critical training and method, however perfect, will
avail him little while he persists in this attitude ; and he will
incur the reproach of deliberately relying upon chance survivals.
I
NON-REPRESENTATIONAL CHURCH ART 131
The specialist will not let himself be blinded in this manner
by the fortuitous.
We have already seen that Islam spread like wild-fire over
the East and attained its second creative centre in the distant
north-eastern corner of I ran . Mecca indeed remained the religious
centre, but apart from this the region of Altai -I ran became the
source of moral and intellectual life. Is it conceivable that Chris-
tianity, which at an earlier date took its origin at no great distance
from Mecca and Medina, can have failed to make its way east-
wards along the same natural path from the Syro-Egyptian end ?
Why does it never occur to archaeologists to include the east
in their survey .'' and why, inspired by the holy fire of the twin
orthodoxies of classical learning and the Church, must they
needs go on to attack the disturber of their peace, who dares,
for example, to ascribe to Mesopotamia a leading role in the rise
of Christian vaulted architecture ? Will they now show a like
hostility to the idea that domed architecture originated in Iran
and travelled by way of Armenia } Arabs and Jews were alike
connected with Iran, the former through the caravan route
across the desert, the latter as a result of the dispersion of the
twelve tribes. The Jew, like the Beduin, found in Iran an atmo-
sphere more suited to his genius and the pure expression of his
non-representational form of religion than in the Mediterranean
region. It was therefore probably due in the first instance to
the Christians of Jewish birth, as in Islam to the Arabs, that
Christian art at its very commencement acquired a non-repre-
sentational character, and subsequently found, like Islam, that
its true and natural support lay in the East. Evidence of this
relationship is still clearly perceptible in Armenia, owing to the
twofold circumstance that the Parthian Arsacids adopted Chris-
tianity as the state religion, and that the most distinguished
of the succeeding dynasties, that of the Bagratids, endeavoured
to prove itself of Jewish descent.^
* As regards the relations of the Jews in Erdkunde, x. The Talmud also
with Armenia and Iran, Leo Meyer mentions Jewish-Armenian learning ; a
writes as follows, with a reference to Jews' high school existed in Nisibis ;
D. H. Miiller, Semitica II, and V. Apto- doctores iudaei are mentioned, and were
witzer. Die mosaische Rezeption im not without influence, as is shown by
armenischen Recht : ' The Jewish popula- the code of the Meckitar Gosh and others,
tion of Armenia was numerous and Felix Lazarus, in Briiks Jahrbiicher, x,
anciently established, as recorded by tries to prove that the legend of the
SchiJrer, Jiidische Geschichte, and Ritter Bagratids is not devoid of an historical
K2
132 NON-REPRESENTATIONAL CHURCH ART, AND
We have seen that as early as the eighth century, when
the seat of Mohammedan power shifted from the Syrian capital
of the Ummayads to the Abbasid capital of Baghdad, the Iranian
spirit became dominant in wall decoration, and the expansion
of its peculiar style can be traced as far as Egypt. Iran in fact
possessed an indigenous art, quite independent of Greece and
Rome, and more closely related to the genius of Islam and of
the Christians of Armenia than to the art of the Mediterranean
area. Is it conceivable that this art should have failed to manifest
itself in the architecture or ornamentation of Christian churches
in the Mediterranean area in the same degree as the various
forms of the vault, viz. the apse, the barrel vault, and the dome ?
Having once established the fact that Christianity developed
on a wide basis on Persian soil and created indigenous forms of
architecture uninfluenced by the Mediterranean, ought we not
to consider the probability that the Eastern spirit had no less
influence in Church-decoration as well ? We shall therefore
inquire whether it is not possible to detect in the barrel-vaults,
domes, and vaulted apses of Early Christian buildings a new
tendency, the mediaeval, foreign to Greece and Rome, and first
borne into the Mediterranean upon the tide of Christianity. The
permanent basis of its strength lay in its markedly Iranian origin ;
it was the same source which ultimately controlled the art of
Islam. This faith, after its brief interlude in Syria, attached
itself to Iran and its culture even as the Christianity of Greece and
Rome embraced a Hellenism already in decline. Between these
two cultural regions were the theological school of Nisibis and its
sphere of influence, both under Aramaean direction, and Armenia,
of which the art was permeated by the spirit of Iran. Originally
all Christian communities maintained constant relations with each
other. It was the conflicting influences set up by the Church, and
especially by the Councils of the fifth century, which first produced
schism and separation, the rise of race against race and nation
against nation, despite the common bond of their Christian faith.
The Council of Ephesus in a. d. 431, insisting upon the divine
motherhood of Mary, provoked the counter-movement of the Nes-
torians in Mesopotamia. The Council of Chalcedon in a. d. 451
was the ultimate cause of the final defection of Egypt and of the
background ; it appears in fact that the David who led their countrymen during
names of the Bagratids occur identically their exile.'
in the lista of the Jews of the house of
THE ANTI-REPRESENTATIONAL MOVEMENT 133
Armenian Church, which had by artful devices been brought almost
to a state of unity in the service of pictures. But while the Semites
reverted to representational art, the Armenian people entered upon
a fierce struggle in defence of their ancient non-representational
style. This internal enmity of the various national churches
increased the existing antagonism of Byzantium and Rome, and had
the effect not only of impeding their own political development
but also of ensuring the subsequent triumph of Islam.
In order to demonstrate the original non-representational
character of Christian art, I shall begin with the decoration of
such vaulted buildings as have survived on Italian soil and are
therefore best known. I shall deal first with the dome, then with
the barrel- vault, and finally with the apse.
I. Italian Mosaics
The method of covering interiors with small glass cubes
probably originated on curved surfaces, to which it is admirably
adapted. In any case it is a fact that strong and creative work
has survived only on vaults, whereas on flat surfaces we find
imitation of well-known forms borrowed from other arts, more
especially from that of manuscript illumination, whether on
papyrus or parchment ; an excellent example may be seen on
the nave walls of S. Maria Maggiore. In seeking examples
of forms adapted to the peculiar character of mosaic it will be
well to confine our attention for the moment to vaulted surfaces.
A. The Decoration oj the Dome. The mosaics of S. Costanza,
in spite of certain Hellenistic intrusive elements, afford good early
examples of the Iranian style. We find here the characteristic
difference between dome and barrel-vault decoration ; I shall
begin with a comparative examination of the former. The
dome was divided radially by richly branching formal tree designs
springing from a river with rocky banks ; a middle zone was
filled with a series of figure subjects. Were we to substitute
Buddhist figures, we should have the decoration of the Naga
cave of SorCuq in Chinese Turkestan (Fig. 42, facing p. 137) ; if
we eliminate the representational element we are left with the
Iranian decorative style, which was the origin of both Christian
and Buddhist roof paintings, though each of these made the
additions natural to its own genms. We find Constantine
Porphyrogennetos still employing the same style of decoration
134 NON-REPRESENTATIONAL CHURCH ART, AND
for the Chrysotriclinion in the palace at Constantinople. In
Armenia the radial division of dome decoration became a per-
manent feature. At Mastara, in the Hripsimeh, at Thalin, and in
other places, we find eight ribs meeting at the top in a medallion.
In some cases the effect was possibly enhanced by painting.
This disposition was carried by the Goths as far as Spain.
More than a century elapsed between the building of this
church under Constantine and that of the Baptistery at Ravenna.
The Hellenistic modification had by now disappeared, not so
the Syrian ; the Aramaean theologians perceived the value of
the radial division of the dome and used it to further their
didactic system. The general scheme of decoration both in
mosaic and stucco remains purely Iranian in form. The interior
of the dome is divided radially into compartments by formal
tree-designs, purely geometrical in structure, and connected at
the top by a border of * lambrequins '. The middle zone below
is filled by a blind arcade in stucco with the peculiar crowning
with which the East-Iranian buildings of Kashmir have
rendered us familiar. The spaces of the arcading round the
octagon at the base are filled with a scroll-design enclosing
central motives, which correspond to the rosettes of the Mshatta
facade.
S. Giovanni in fonte in Naples occupies an intermediate
position between S. Costanza and the Baptistery of Ravenna.
Its Iranian character is at once betrayed by its square ground-plan
and by the use of squinches to effect the transition from square
plan to dome. The interior of the dome is once more radially
divided by formal trees connected at the top by ' lambrequins ' ; in
the compartments formed by this ornament and in the spandrels
above the squinches are characteristic Hvarenah motives in
the form of a vase between birds, or of whole landscapes with
sheep or stags by the side of a shepherd and flanked by palms.
A special group is formed by the domes of the two cruci-
form churches with a single nave, namely, the tomb of Galla
Placidia and Casaranello. Here we see the cross depicted on
a starry sky as background. These features seem to indicate an
Eastern origin, and the supposition is confirmed by the decoration
of the adjacent barrel- vaults. I shall revert to this question
when discussing the apses.
B. Barrel-vault Decoration. In this group too we can begin
with the mosaics of S. Costanza, though now with those of the
THE ANTI-REPRESENTATIONAL MOVEMENT 135
circular aisle. The barrel- vaulting of this aisle is divided into
twelve distinct panels, which are not homogeneously treated ;
each is filled independently with a continuous pattern, but the
arrangement is such that opposite pairs have identical designs,
of which there are thus only six. One of them consists of a vine-
scroll formally treated, though certain Hellenistic features show
that it is intended to suggest the vintage. Another shows a
familiar Iranian motive introduced through Syria, detached
branches of pomegranate type, with a bird or a vase to almost
every branch. One of the barrel-vaults of Quseir 'Amra has the
same motive, but in a lozenge-diaper. It is very common in
the Eusebian canons of Syrian and early Armenian Gospels,
and reappears in the Mesopotamian Gospels of Rabula (a. d. 586)
as a decorative detail near the base of the arcades. Two other
corresponding vaults in S. Costanza are decorated with inter-
connected circles ; in one case these are triple-banded with
contrasting colour. The last two pairs have a four-rayed and
a six-rayed pattern of the kind which reappears in Amida and
Egypt. The six-rayed variety was afterwards very popular
in Islamic polygonal designs.
Equally rich examples of barrel-vault decoration have been
preserved in Salonika. In the Church of S. George remains of
mosaics were saved from destruction by Rosi in 1889 ; with the
exception of one piece, these might quite well be imitations of
Persian silk fabrics, as might be those of S. Sophia in Constanti-
nople. In front of the apse in the Church of S. Sophia in Salonika
there is a barrel-vault decorated with a cross within a circle on
a gold ground, and bordered with a textile repeat-pattern com-
posed of vine-leaves and crosses. In the West similar barrel-
vaults occur in the nave of the mausoleum of Galla Placidia
at Ravenna. There the transept is decorated with gold vine-
scrolls enclosing the sacred monogram on a blue ground, while
the window is surrounded by acanthus scrolls and figures of stags
at watersprings. The latter designs are enriched with colours
suggested by bird's plumage, and differ from those of S. Costanza,
which have a white background. I shall return to this subject
later.
Another type of barrel-vault decoration is described by
Agnellus as occurring in the Basilica Ursiana at Ravenna. The
Syrian Bishop Ursus had the testudo of this church decorated with
mosaics before the year 400, the work on the women's side being
136 NON-REPRESENTATIONAL CHURCH ART, AND
executed by the mosaicists Eusebius and Paulus, that on the men's
side by Satius and Stephanus. The passage almost certainly
refers to vault decoration : et hinc atque ulinc gipseis metallis
diver sa hominum, animaliumque et quadrupedum enigmata inciserunt
et valde optime composiierunt. In a letter written by Nilus of
Sinai a few years later there is a reference to the use of stucco
in church decoration as a means of pleasing the eye with figures
of creatures flying, walking, and creeping, and of every kind of
plant. We are reminded of the use of Hvarenah symbols in the
decoration of Mazdean interiors ; there we find a combination
of similar motives in stucco, the function of which is not to please
the eye or serve as aenigmattty but rather to encompass the room
with the glory of God. A good example of this type of decoration,
modified of course by Christian influence, is preserved on the
chancel roof in the Church of S. Vitale at Ravenna. It consists
of four formal tree-designs converging from the corners to
a crowning circle, the intermediate spaces being filled with
continuous scrolls enclosing a large number of birds and animals.
C. Apse Decoration. There are two apses in Rome which
have always attracted notice on account of the river-landscapes
on their lower borders. Unhappily they were both altered about
A. D. 1300 ; but there is little doubt that in one of these buildings,
S. Giovanni in I^aterano, the existing decoration has preserved
the original cross above the hill with the rivers of Paradise ;
in the other building, the Church of S. Maria Maggiore, the
lateral scroll-work with its figures of birds and animals is almost
V certainly part of the original decoration. Thus a combination
of the features exhibited by these two apses would show a river
landscape, surmounted in the middle by a cross, and bordered
by continuous scrolls enclosing animal figures as a surface decora-
tion. Whence could so curious a composition have originated ?
Other similar apses exist in a fairly good state of preserva-
tion, but in these the river-landscape at the lower border is no
longer to be seen. Thus in the vestibule of the Lateran Bap-
tistery (SS. Seconda and Rufina) we find a purely formal per-
pendicular treatment of the middle portion ; in S. Clemente,
a building of later date, the cross has been superseded by the
crucifixion, although the design of birds and animals retains its
full luxuriance to delight the worshippers' eyes. These examples
suggest that the cross was not originally an integral part of river
landscape and scroll designs.
b
/
X-.,
m.
:=ew^>:'^
mj^
^. (l-i=r-'
1
42 Sorcuq, Chinese Turkestan ; fresco in the Naga
Cave ; after A. Griimvedel. See pp. 120, 133.
r
^ 4
^^^H
^^^^^^^\
^^^^^^^^^^^^^^^^^^^^^^HH^^^r* sl^^F^^^^^^^
* ^^^^^^^^^^^^^^^^B
43 Oasis of EI-Khargeh ; apse decoration of a funerary chapel ;
after V. de Bock. See p. 137.
THE ANTI-REPRESENTATIONAL MOVEMENT 137
In the Church of S. ApolHnare in Classe another type of
apse decoration has survived from the sixth century ; its existence
as early as the years a. d. 401-3 is attested by Paulinus of Nola.
It consists of a landscape into which symbolical figures of a lamb
and a dove are introduced, and over which a cross is suspended.
It is important to remember that such landscapes are placed
in the apse, the point on which the eye of the spectator is focussed ;
we have therefore to deal with art on a monumental scale.
The most important remaining example is the apse of S. Apol-
Hnare in Classe, the lateness of which is always felt to be an
anachronism. In type it really belongs to the fourth century ;
thus the way in which the sun's disk, here transformed into the
cross of Golgotha, floats large over the landscape, almost makes
us forget to read the symbolism of the Transfiguration.
Examples of such landscape also occur in other churches in
all the various parts of the nave. Thus we find the Good
Shepherd represented over the western door of the mausoleum
of Galla Placidia, and as a recurrent motive on the nave- walls
of S. Maria Maggiore ; a pastoral landscape recurs again and
again in the scenes from the Old Testament. In the same
position in the Church of S. Sergius at Gaza was a representation,
praised by Chorikios, of waters enlivened by birds among
flowering meadows, and S. Nilus even alludes to hunting and
fishing scenes. Evidently these scenes form a distinct group, in
which the two river landscapes of the above-mentioned apses
should be included. The importance of this kind of landscape
decoration in the earliest churches is strongly emphasized by the
fact that the iconoclasts revived it, a circumstance which seems
to indicate that it can hardly have been confined to the surviving
examples in Italy, but must have had its original area of distribu-
tion in the East.
There are isolated examples of apses which make no attempt
at representation ; instead they contain purely geometrical orna-
ment, or, at most, rows of symbolical devices. One of these still
exists in the oasis of El Khargeh (Fig. 43, opposite); here
eight-pointed stars connect octagons each filled by four hearts,
and these form a repeat-pattern. Similar designs are commonly
used to fill the arches of doors.
In this survey of evidence for an Early Christian art of
a non-representational character, we are immediately struck by
its wide distribution in the fourth century ; whereas only a few
138 NON-REPRESENTATIONAL CHURCH ART, AND
straggling examples can be adduced for the later period, until
its subsequent revival on a large scale by iconoclastic movements.
It appears in fact to have been completely ousted by representa-
tional art. All the examples quoted belong to the Mediterranean
sphere of influence. What then was the attitude of the Christian
East ? It might seem obvious to make responsible for the non-
representational style the people among whom the Christian
religion had its origin. We are naturally tempted to think of
so-called ' Christian Art ' as directly inspired by the Founder.
But do we in fact find the slightest evidence for associating
Christ with an artistic movement of this nature .'' Christ resembled
Buddha, Zarathustra, and Mohammed in excluding art from
the activities indispensable to the religious life. The attitude of
Buddha and Zarathustra towards representational art is sufficiently
explained by their Aryan origin ; Christ and Mohammed be-
longed to branches of the Semitic race which, unlike the Mesopo-
tamian Semites of the Great Monarchies, never represented in
art. The explanation must be sought in the fact that they were
pastoral nomads at the time when their religions were founded.
If, then, the Founders ignored representational art as a
means of furthering their aims, who was the first to make use of
it ? Certainly neither the apostles nor their immediate successors.
Moreover, the satisfaction of definite individual needs led at
first to forms dictated less by art than by utility. I cannot but
conclude that the character of Christian art, as of architecture,
was determined in the first instance by the several nations among
which Christianity struck root, and in the next by those great
regions of East-Aryan and West-Aryan culture into which
it subsequently expanded. Such is the path of inquiry I shall
pursue, commencing with the middle regions.
II. The Eastern * Hinterlands * of the Mediterranean
The origin of a certain kind of non-representational sym-
bolism can perhaps be attributed to the Jews. I refer to the
embodiment of the divine in the form of a lamb. ' Lamb of
God ' (John i. 20) would be a metaphorical expression natural
enough to a people of shepherds like the Jews. Equally natural
would be the representation of the apostles as lambs, after
the fashion of the apse-mosaics already cited, and of many other
examples. The same origin is perhaps even more definitely
THE ANTI-REPRESENTATIONAL MOVEMENT 139
suggested by the introduction of whole scenes in the spandrels
of the blind arcades on the sarcophagus of Junius Bassus or on
the Ravenna sarcophagi. Even a decree of the Council of 692
was unable to suppress this deeply-rooted symbol. It was not
indeed employed in Jewish synagogues ; but in them we do at
least find an instance of plant symbolism in the palm.
Other examples from Jerusalem and the Syrian, coast are
certainly not indigenous, but belong to the hybrid art of the
Mediterranean, though showing in part an increase of oriental
influence, as will appear below.
In the church architecture of East-Syria, the period of which
certainly falls between the fourth and the beginning of the
seventh centuries, representation of the human figure is entirely
absent. In this region, curiously enough, the earliest representa-
tional paintings known to exist are those of Quseir 'Amra, be-
longing to the Ummayad period. In the main part of the building
are hunting and other scenes which may be classed with land-
scapes, while the roof decoration recalls the Indian paintings
of Ajanta. The ornamentation of the Baths in this desert palace
is rather Indo-hellenistic in style, with the exception of a diaper
enclosing animal figures on one of the barrel-vaults. In Christian
East-Syria we shall search in vain for anything beyond symbols
on doorposts, such as the cross, or stars enclosed in circles, and
a few other very modest ornaments.
In Mesopotamia, again, paintings have not been discovered
as yet in wide distribution. This country is the home of the
barrel- vaulted church with transverse nave, and of the single
long nave with barrel-vault ; on its borders we also find the type
with barrel -vaulted nave and aisles, the best examples of which
have survived in Armenia and in the interior of Asia Minor.
As in the case of the dome, we are confronted with the question
whether these barrel- vaults and apses were, or were not, originally
destitute of ornament. In the Tur 'Abdin the apses of a number
of churches have a cross in low relief, and the apse of the Basilica
of Resafa still contains original stucco-decoration, with radiating
bands connected at their bases by arches. Might we not reason-
ably expect to find similar ornament elsewhere in stucco, mosaic,
or painting ?
Quite distinctive, in this region, is the decoration of the
architectural members adjoining the curved surfaces. Our
knowledge is at present limited to the triangle of towns in the
140 NON-REPRESENTATIONAL CHURCH ART, AND
North : Edessa, Nisibis, and Amida. There we find that the
Greek architectural members, designed for organic construction,
have been adapted to a system in which decorative Hning was
all-important. The transition can hardly have originated in
stone churches, such as those which have survived ; it probably
first occurred on buildings of unburned brick, and therefore
in Iran or South Mesopotamia. Gandhara, like Armenia, shows
no trace of it. I should, therefore, be inclined to rule out Eastern
Iran as a possible source. There the simple interfacings and
repeat-patterns of the kind proper to wall-lining are indigenous,
and entirely unaff'ected by classical influence. But on the lower ^
Tigris and Euphrates, such a transition from the constructional ;
to the decorative may quite well have occurred. In the present '
state of our knowledge we can only infer back from the stone
buildings which have survived ; and these have in all probability
been affected by Iranian as well as Mesopotamian influences.
The monuments are mentioned in my work on Amida. The
chronicle of Soort allows us to conclude that there was a connexion
with Antioch.
In addition to the general spread of Hellenism under the
Seleucids, there was a special circumstance affecting South
Mesopotamia which cannot be left out of account. I refer to
the transportation of the inhabitants of Antioch after it had been
twice conquered by Shapur in a.d. 256 and a.d. 260. The captives
were settled in certain towns in Babylonia, Susiana, and Persis.
' They were given allotments of land to cultivate and houses to
' dwell in. In consequence Christians became numerous in the
* Persian Empire, and monasteries and churches were erected.
* There were priests among them, who had been carried off as
* captives from Antioch ; they dwelt in Gundisabur, and elected
' Ardak of Antioch bishop, since Demetrius, the patriarch, had
' fallen sick and died of grief The Christians spread throughout
* the empire, and their numbers multiplied in the East. In
' Rasahr, the seat of the Archbishop of Persis, two churches
* were built, one of which was named the Church of the Romans,
' the other the Church of the Syrians. The service in them was
* conducted in the Greek and the Syrian tongues.' ^
In my opinion it was this current of Antiochene influence
coming from South Persia which in the North brought about
^ I have taken these passages of the Soort Chronicle from the translation of Sachau.
THE ANTI-REPRESENTATIONAL MOVEMENT 141
the obvious transition in stone buildings, from the Greek to the
Perso-Mesopotamian style. This would appear to have been
the result of a return migration from the South and not of any
direct impulse from Antioch. This is borne out by the appear-
ance of Assyro-Babylonian features in Greek forms of ornament,
an instance of which is the breaking-up of the frieze. On the
other hand the shapes of individual capitals and their decoration
are Iranian.
Armenia was the only point at which the West- and East-
Aryans of the South came into direct contact ; and there
Christianity, adopted as the official religion about A. D. 300,
preferred Mazdean to Hellenistic forms. The oldest surviving
Armenian churches date back to the fifth century ; inscriptions
prove their existence in the sixth century, and their wide distribu-
tion in the seventh. In the matter of exterior decoration a number
of churches show a rich and homogeneous style ; examples
occur in the case of the Cathedral of Artik (an apse-buttressed
square), and that of Thalin (a trefoil-ended church with three
aisles), at Zwarthnotz (a quatrefoil with ambulatory), and Irind,
which has an eight-foil plan. The decoration consists of blind
arcading with cushion capitals, the arches themselves being
ornamented with interlacing of three-grooved bands, or with
vine and pomegranate foliage. In addition there are arcaded
fillets or slant surfaces with interlacing under the roof, and deco-
rated bands following the curves of the window heads. Of the
interior decoration nothing remains ; but documentary evidence
goes to prove that Armenia, so far from being an original home
of representational art, received it first from the Greeks, while
it became one of the chief centres of hostility to pictures.
The question now arises whether the vaulting of Armenian
domed buildings was devoid of all coloured decoration. I have
already mentioned that in the domes we often find eight ribs
radiating downwards from the Crown and terminating in disks.
This would seem to imply the absence of figure art ; but does it
justify us in ruling out every other kind of decoration in painting,
mosaic, stucco, or other material ?
In Armenia we find ourselves surrounded by one of the
main groups of vaulted buildings, the unit with a single central
dome ; this in its early development is accompanied by apsidal
buttresses ; soon afterwards, barrel-vaults are interposed between
these buttresses and the central bay. Are we to suppose that all
142 NON-REPRESENTATIONAL CHURCH ART, AND
these vaulted surfaces were entirely destitute of ornament ?
In the oldest surviving buildings in Armenia, dating from the
fifth to the seventh centuries, we now find no contemporary
figure-paintings ; for the original decoration we can at present
only adduce documentary evidence. Peculiar importance attaches
to a statement occurring in Moses of Kaghankatuk's history of
the Albanians, which dates from the seventh century. He
narrates that, while the Armenians and Byzantines were engaged
in disputes and rivalry, peace reigned among the Albanians
until news reached them of a movement hostile to pictures.
Thereupon Bishop David referred the question of paintings
and drawings to John Mairagomier and received the following
instruction : ' This sect (those in favour of ikons) arose after
' the time of the apostles, first appearing among the East-Romans ;
' wherefore a great Synod was held in Caesarea, and the painting
' of pictures in the House of God was sanctioned. Hence the
' painters grew arrogant, and wished to place their art above
' all the other church arts.' They said : * Our art is light, since
* it is the means of enlightening old and young alike ; whereas
* but few can read the holy scriptures.' Hence arose that con-
troversy which spread to Albania, and involved not only painters
and writers, but whole sects of iconoclasts and ikon-worshippers
in conflict.
In Armenia the defeat of the painters was a foregone con-
clusion, both on account of the ingrained opposition to repre-
sentational art in that country and because the Armenians had
a script of their own in addition to the ecclesiastical language.
A further reason was the embittered conflict with Byzantium,
and the separation of the Armenian church, which repudiated
the decrees of the Council of Chalcedon. This Council (a. d. 451)
had far-reaching effects upon the history of art. The mono-
phy sites, and with them the Armenians, adopted a hostile attitude,
and soon regarded all who accepted the decrees as their bitterest
enemies. At the Council held at Vagharshapat in 491 the
decrees of Chalcedon were formally condemned ; as Topdschian
justly observes : * The Armenians sacrificed everything in order
' to protect the foundations of their national Church, which
' has to the present day proved an important religious and
' political factor in the preservation of their nationality.* It is
hardly possible that an entry into Armenia could have been
effected before the Byzantine period (i. e. before the tenth to
THE ANTI-REPRESENTATIONAL MOVEMENT 143
the eleventh centuries), by any iconography expressing the decrees
of the Council of Chalcedon as described by Corippus, pre-
sumably in connexion with the confirmatory Council of A. d. 553
in the Church of S. Sophia. True, there was a period of eighty- \
nine years — between the Council of Karin in a. D. 630 and that t> V",
of Manazkert in a. d. 719 — during which six patriarchs (including \ f^
Nerses III the constructor) followed the * shameful doctrine of J
Chalcedon ', until John_the philosopher (a. d. 717-28) re-
_established monophysitism. But as he simultaneously opposed the
Paulician sect, his period is one of considerable unsettlement in this
matter. Did the remains of mosaics and representational paintings
found in ruins, together with isolated mention by historians,
convince us that decoration with cycles of representational paint-
ings was habitual in these early Christian churches, then we should
have to place Armenia in the same class as the Mediterranean
area and the West. But it is precisely here, I believe, that one
of the characteristic contrasts between Armenia and that more
familiar culture of the West is most conspicuously displayed.
Ter-Mkrttchian points out that in the twelfth and thirteenth
centuries the Armenian patriarchs defended themselves against
the reproaches of the Byzantine Emperor for refusing to accept
sacred painting, while in the tenth patriarch Wahan (a. d. 968-70)
was dismissed and condemned for introducing them and for
banishing the glory of the cross from all the altars by substitution
of ' ikons '. It is not likely, he argues, that Armenia's attitude
towards sacred painting in the sixth and seventh centuries was
any more favourable than in these later periods. ' They probably
' never had them in their church and had no particular reason
' for discussing the matter. The question was first brought to
' a head through the attacks of the Greeks ; and among the
' iconoclasts of the time of Moses (a. d. 574-604) arose a party
' which opposed the veneration of pictures as an unchristian
' innovation, and, perhaps influenced by existing sects, extended
' their hostility to all forms of religious expression. This hostility
' inevitably gathered strength and intensity in proportion as
' hatred of the Byzantine Greeks increased with the growth
' and spread of their cult of pictures. It may be laid down as
' a general rule that figure representation in church interiors is
' a sign of Aramaean or Greek influence.' In Armenia we can
see clearly enough the positions taken up in the struggle by the
parties which arose at the change from the nationalism of the
t
144 NON-REPRESENTATIONAL CHURCH ART, AND
fourth, to the ecclesiasticism of the fifth century. But supple-
mentary evidence of great value is furnished by a document
actually written during the transitional period ; this was the letter
sent by Nilus of Sinai to the prefect Olympiodorus at the beginning
of the fifth century. It commences as follows : ' Having under-
taken to erect a large temple in honour of the holy martyrs,
you ask me whether it be proper and fitting, in the first place,
to introduce figures of them in the choir illustrating the toils
and sufferings wherewith they testified to their Christian faith
in the agony of death ; and secondly whether it be proper to
cover the walls with all manner of hunting scenes both upon
the right hand and upon the left, depicting snares spread out
upon the land, and likewise hares, deer, and other creatures in
flight, also figures of men pursuing them breathlessly with their
dogs, desiring to slay them ; or again, nets let down into
the sea and fifled with all kinds of fish, or drawn up on the dry
land by the hands of the fishermen. Further, whether it be
right to introduce all manner of stucco ornament, a pleasure to
behold in the Lord's house ; and yet again whether you should
decorate the part used for the congregation (accessible to all the
laity) with thousands of crosses, and representations of flying,
walking, and creeping beasts and every kind of plant.'
In the continuation of this letter Nilus denounces this kind
of ornamentation with such vigour that he is obviously dealing
with a wider question than that raised by the prefect. In point
of fact the kind of decoration referred to here is identical with
that which reappears later both in Byzantium and farther East
during the iconoclastic period, so that we are justified in speaking
of it as a kind of reversion to the original East-Christian art.
I shall now pursue its traces in the region which lies beyond
the interior region east of the Mediterranean, bounded, roughly
speaking, by the river Tigris.
III. The East-Aryan Province. In this section I shall be
compelled to depend on presupposition, first derived from the
evidence of Armenia, and secondly from monuments both in
East and West, which have preserved in a durable material the
perishable decoration of original Iranian buildings. Armenia
was at first very poor in architectural ornament ; it borrowed
r from Iran the blind arcade and the manner of filling the spandrels
\ with geometric forms of vine foliage, pomegranates, animals,
» and birds. Zwarthnotz (a. d. 650) and the Church of Gregory
THE ANTI-REPRESENTATIONAL MOVEMENT 145
of Honentz in Ani (a. d. 1215) represent the average type;
Achthamar (a. d. 915-21) with its rows of animal figures and its
frieze depicting the chase shows the pure unmodified Iranian
style. I shall now give a systematic review of this group of
buildings according to their different styles of decoration.
A. Structural Members. In Mesopotamia we found merely
a modified form of classicism in the structural members ; the
evidence relating to Iran is altogether different in character.
{a) The Support. The evolution of this member is most
important both in a geographical and a chronological sense ;
it serves as a criterion by which to test the progressive fusion
in Early Christian architecture of the ' organic ' Greek style
and the decorative style of Iran, Assuming that classical antiquity
was acquainted only with the support and the architrave, and
Iran only with the wall and the arch, we find the first attempt at
fusion in the combination of the column or the pier with the arch.
Whether this first occurred in Christian times or earlier is irrele-
vant. The combination was certainly effected at an early date
in isolated examples. In the indigenous art of East Syria it was
confined to pier and arch. In the palace of Diocletian {c. A. D. 300),
which bears clear traces of Syrian influence, and in Syria itself
after the fourth and fifth centuries, the process was freely extended
to the column and arch.
The migration of the vault and the arch from East to West
was accompanied by quite new forms of decoration. We have
already observed these in their Mazdean garb and have now to
inquire how Christian architects gradually adapted themselves
to the new style. This can best be done by examining the
motives of engaged shaft and column.
The national art of Iran, if we exclude imperial palaces,
was unacquainted with the column, which in classical architecture
was used in definite proportions of diameter and height as a
detached perpendicular support for a pressure horizontally
distributed. The vaulted architecture of Iran was only acquainted
with the engaged shaft, a projection running up the wall and
dividing it perpendicularly into compartments. The engaged
shaft quickly became acclimatized in Christian Armenia and in the
last quarter of the first millennium it coalesced with the pilaster to
produce the clustered column (Figs. 21 and 23, facing pp. 94 and
95), which is so familiar and typical a feature of Western, and more
especially of Gothic architecture at the beginning of the second
a4S» L
146 NON-REPRESENTATIONAL CHURCH ART, AND
millennium. Judging by its distribution in Armenia, India, and
the Islamic countries, the engaged shaft in Iran must have had two
forms of termination, the 'knob ' and the cushion capital, both
of them probably derived from Aryan wooden architecture. That
explains how it is that we find the cushion capital, which entered
Lombardy probably under Armenian influence (Figs. 2 and 24,
facing pp. 62 and 95 ), appearing simultaneously as an indigenous
feature in the North. The columns of the Norwegian wooden
Church of Urnas, if pre-Romanesque, illustrate its wooden
origin. The ' knob ' termination was adopted by preference in
Islam as soon as it began to develop an individual, i. e. Iranian,
style, distinct from that of the Mediterranean.
The decoration of the ' knob ' and cushion capitals is never
derived from plant life ; or, if so, it is purely a surface pattern,
and never suggests the freedom of organic growth. On the
contrary the shaped sides of the ' cushion ', which occurs both in
Armenia and the West, suggest something hanging down.
And in point of fact they probably have some connexion with
textile hangings.
More important than Islam or the North for our present
purpose is the district surrounding the islands of Marmara
at the gates of Constantinople, The ancient marble quarries of
Cyzicus acquired a new value when Constantine designedly
founded his new city in their vicinity and collected his workmen
from every corner of the Empire. The presence of Armenians
and Iranians among them is proved by the line of development
followed by the capital.
At first its old composite form is retained, and only the
acanthus is modified in the direction, apparently, of Anatolian
style. As we know from the Sarcophagi — of which the * Christ '
relief in Berlin is a good example — a favourite practice in the
* Cilician corner ' was the production of deep shadow effects by
-drilling ; the same device appears in the Theodosian capitals,
which were sold and exported in all directions from Constanti-
nople. This modification of the leaf was followed by another
affecting both the shape and decoration of the capital, which,
if it came from the Mediterranean, would be quite unintelligible.
This new type, which is intermediate between the cushion and
the knob, I have called the impost capital ; in it the square sides
are gradually rounded oft" for purely technical reasons. Even
if there were any doubt as to the origin of its form, the decoration
WM
44 Angora, Asia Minor :
so-called column of Augus-
tus. See p. 147.
45-47 Bisutun and Ispahan
capitals. See pp. 114, 147
48 Edessa ; impost capital.
See p. 147.
49 Baghdad ; impost capital.
See p. 147.
50 Meiafarqin, Church ot the \ irgin ; interior from West door ;
photo., G. L. Bell. See p. 147.
^l0[^i^^ "■« '
51 Mshatta, capital fromthe trefoil-ended hall, 52 Dara ; basket capital.
Trikonchos ; left half. See p. 147. See p. 147.
THE ANTI-REPRESENTATIONAL MOVEMENT 147
of this capital stamps it beyond question as distinctively Iranian.
Each side is as far as possible treated as an independent panel,
and filled with geometric interlaced or scrollwork patterns. I
have collected a few examples from Angora, Bisutun, Ispahan,
Edessa, and Baghdad (Figs. 44-9, facing p. 146), and assume that
the reader is familiar with those from Constantinople. The space
between the square abacus and the springing of the arch is
frequently filled by an impost block, which like the capital itself
is covered with sacred monograms or unequivocal symbols of
the glory of God. The ' knob ' is transformed at times into the
representation of a basket after the Greek manner, a change
suggested both by its rounded form and interlaced ornament.
Again I may quote a few examples from the East (Figs. 50-2,
opposite), viz. the Church of S. Mary at Meiafarqin, Mshatta,
and Dara. This ' basket ' is sometimes wreathed with a design of
vine-leaves and bordered with the Hvarenah animals of ancient
Iran, rams, or goats, birds, and lions. In Armenia capitals with
rams' heads called chojak were known in the fourth century,
reminding us of the Acnaemenian bull capitals ; in the Cathedral
of Zwarthnotz we still find the basket form of capital with Ionic
volutes. The usual type was the knob surmounted by interlacings
and by circles bearing symbolic emblems.
{b) Door and Gate. The East-Aryan treatment of the door
differs from the West-Aryan. Its distinguishing features are
the trilateral or round termination above and the porch, both of
which are derived from timber buildings. The wooden churches
of the Ukraine contain examples in the original material ; older
still are the stone imitations in Kashmir of the second half of
the first millennium, while the churches of Armenia illustrate
its full development in rubble concrete. There, too, we can
clearly trace how the fusion of the round door and the porch
produced the recessed door, which later became acclimatized
in the West, and developed on so magnificent a scale. This
fusion was effected by a series of recessed colonnettes ; the idea,
achieved only bv degrees, was to make those who entered feel
that they were being drawn through a kind of funnel into the
interior of the building. In Armenia, where this development is
well illustrated, it is carried to further lengths than in the West,
where possibly it had a more or less independent origin in wooden
architecture. The method adopted was to amplify the arch by
surmounting the original semicircle with others rising pro-
L 2
148 NON-REPRESENTATIONAL CHURCH ART, AND
gressively higher in the centre, until finally the largest assumed
the shape of a horseshoe (Fig. 53, facing p. 152). In the West
the addition of human figures, which were naturally absent in
Armenia, led to that richness of effect which eclipsed even the
' High Gate ' of Seljuk mosques. Nevertheless comparison
between the two is not unprofitable. In both cases the more
showy side of church architecture was concentrated exclusively
upon this member with its accessories ; the latter took the form
of lateral porches in the one case and walls with fountains in
the other, to which were added towers serving respectively for
bells or for the muezzin. Observations of this kind may
ultimately stimulate comparative research in art, which can
hardly fail to produce better results than the existing historical
and aesthetic classical s^l, ;n'f^ with the local limitations of their
narrow horizon. / ^^
B. Decoration oj Jia^ y,urfaces. This subject has already
been discussed in part. Here I shall deal only with a particular
type. It has been shown that a form of symbolical landscape,
strikingly similar to that found in the earliest Christian apses,
was also a feature of Mazdean art. But there is a school of
thought which regards this same landscape-decoration as a sur-
vival of Greek naturalism, and attributes the appearance of
* stiffness ' to the introduction of a gold background. In opposi-
tion to this view I would tentatively put forward the hypothesis
that this estrangement from nature has absolutely no connexion
with the use of the gold background to produce an effect of
unlimited space ; nor can it be regarded as a last degenerate stage
of classical art ; rather it is connected with the spread of influences
from Altai-Iran in the late-Roman period. This landscape is
not based upon observation of nature, but must be regarded as
a congeries of Mazdean symbols forming a composite whole.
Clouds, water, and earth, together with certain symbolical
animals and plants, constitute the main subject-matter, the
terrestrial section being supplemented by the sun, moon, and
stars. It is admittedly impossible at present to point to a single
extant Mazdean work of this kind ; but the art student cannot
afford to ignore this hypothesis as lightly as the dilettante,
despite the rule of reckoning only with concrete evidence
and disregarding the notorious gaps in our record of existing
monuments.
In the course of this inquiry I have been led to assume the
THE ANTI-REPRESENTATIONAL MOVEMENT 149
existence of a counterpart to ' Christian classical art * in ' Chris-
tian Mazdean art ' ; by this I mean the form of Church art
existing in the fourth century, which even in the Hellenistic
area borrowed not only the vault but also its form of decoration
from Mazdaism. The most important examples are dome
and apse mosaics at Rome, so far as they can be assigned
with certainty to the fourth century. These are supplemented
by contemporaneous and later examples in Milan, Naples, and
Ravenna, while there are a few isolated traces in the East. I
shall take the various types and forms of decoration in their order,
commencing with the cross, on account of the way in which
it was frequently made to dominate the whole landscape (see
the letter of Paulinus of StolaV
(a) Crosses. India originally representations of Buddha;
so too in Iran the figure of Chrirf' k/« not represented. Maz-
daism made use of religious symoois ; similarly the Aryans
in India regarded as emblems of divinity the wheel, the tree,
and later the stupa, but not originally the human form. In the
East-Aryan region the Christians universally adopted the cross
as their favourite symbol of Christ ; in fact we might regard
it as one of the three means of depicting the religious Founder,
the Iranian ; the other two were the Greek unbearded type of
Christ, and the Semitic bearded type. We cannot here discuss
the origin of the cross itself ; the vision of Constantine gives food
for reflection. There is no need to quote evidence for its existence
in the most varied forms, more especially in that of the swastika,
in pre-Christian times. Characteristic of the Iranian cross are
its various kinds of terminal additions, e. g. single disks or pairs
of loops. There is no ground for asserting that in the oldest
Roman mosaics its place was taken by actual representations of
our Lord.
(b) Landscapes. I regard the apse of S. Aquilino in Milan,
which has survived from the fourth century, as an example of
a Christian landscape of the Mazdean type. It was actually
possible to discover vestiges of the sun-god depicted as riding
m his chariot above the figures of the shepherds below. He may
also have been accompanied by the moon and stars. Beneath are
those conventional rocks from which on either side water is seen
flowing towards the foreground. To the left are sheep, to the
right cattle with their herdsmen. Here we have a definite
Hvarenah landscape partly modified by Hellenism. We can
150 NON-REPRESENTATIONAL CHURCH ART, AND
still observe on late-Armenian tombstones the manner in which
the sun and moon were depicted in Iran, as disks made to resemble
faces above running animals. The mosaic of S. Aquilino did
not originally stand alone. In the porch of Constantine's Baptis-
tery at the Lateran, opposite the apse decorated with scrollwork
(p. 136), was a second apse with a landscape, in which the Avesta
symbol of the sun had been supplanted by the cross. The
' inappropriate decoration ' of this mosaic was inexplicable to
the classical Christian school. Unfortunately it is now completely
lost ; but records and partial imitations establish beyond doubt
that the subjects depicted were herdsmen with their herds,
others with birds, aviaries, and cleverly executed emblems,
together with trees and flowers.^
(c) Scenes from the Chase. Running round the exterior of
the cruciform church of Achthamar, built on an island of Lake
Van between the years a. D. 915 and 921, there is a median
frieze in low broken relief (Figs. 39 and 57, facing pp. 122 and
159) depicting archers and animals of various kinds in a remark-
able forest consisting of scrolled vines and pomegranates. On
the same frieze also occur figures, not of vintagers at work as on
Roman sarcophagi, but represented frontally in a squatting attitude
after the Persian manner, embracing the vines and grasping at the
clusters or raising their cups. Thus we have here a combined
representation of the joys of the chase and the pleasures of wine.
All this is as little connected with the Christian church as the
hunting and fishing scenes depicted on the interiors of the fourth-
century churches mentioned by Nilus, or the remarkable river-
landscapes which make their appearance in the Roman mosaics.
(d) Vine Scrolls. The origin of the vine scroll and the
remarkable richness of its employment in Christian art is cus-
tomarily accepted without question as adequately explained by
its symbolical and allegorical character. But the student of art
is gradually beginning to take a wider view. He perceives that,
originating in India and Iran, it spread to other countries as
a purely decorative design, being transformed beyond recognition
in Rome by fusion with the acanthus, converted into the palmette
in Islam, and into the lotus in Buddhist countries. The Hvarenah
facade of Mshatta is a comparatively late specimen of this type
^ Compare the well-known minia- Indicopleustes (Repertorium fur Kunst-
tures of the Vatican Virgil and those of wissenschaft, xxxix [1916], p. 243 f.)-
the Christian Topography of Cosmas
THE ANTI-REPRESENTATIONAL MOVEMENT 151
of decoration, but unless its early existence is assumed, the
scroll decoration of the Ara Pacis at Rome and certain variant
forms of the so-called Arabesque are not easily to be explained.
The principal example of Early Christian treatment of vine ■
scrolls in the Mazdean style is the well-known episcopal chair
of Maximianus at Ravenna. The upper and lower borders of
the front are decorated with handsome vine scrolls interspersed
with various animals, including stags, peacocks, and humped
oxen. This design extends over the uprights in a continuous
pattern. Only the panels are adorned with figure subjects ;
we shall have occasion to refer to these below. The question
now confronting us is whether this form of decorative scrollwork
is really Christian in origin, as has hitherto been generally
assumed, or whether it should not be ascribed to Mazdean
influences accompanying the import of ivory from India by way
of Persia. Conclusive evidence is not at present available, but
the constant recurrence of this motive on vault mosaics suggests
its ultimate derivation from a common source. The vine scrolls
in barrel-vaults and apses are constantly associated with the
life-giving water, although the latter is supplanted by the cross
with gradually increasing frequency. A striking example of its
modification for Christian purposes is the symbolic picture of
Psalm xli, viz. a pair of stags drinking from a single spring,
which occurs in the Baptistery at Naples, in the mausoleum of
Galla Placidia, and in several other places. This spring is depicted
issuing from a rock in the manner already described m the land-
scape of S. Aquilino.
The best example of the symbolic treatment of vine scrolls
and palms in mosaic is the roof of the Matrona chapel of S. Prisco
at Capua. Not very different, we may suppose, was the original
appearance in colour of the Mshatta fa9ade, making due allowance
for the diff'erence between the mediocre craftsmanship of a stagnant
art and the brilliant powers of execution possessed by a culture
to which that building is the sole surviving testimony.
(e) Dazvn clouds. One of the most remarkable intrusive
elements in Christian mosaics consists in the highly conventional
clouds which appear as a number of wavy lines in colour super-
imposed on a blue ground. They are purely symbolic in character,
and can in most cases hardly be recognized as pictorial features.
The best example is in the two lateral arches of the Matrona
chapel of S. Prisco, where the Glory of God arises in the centre
152 NON-REPRESENTATIONAL CHURCH ART, AND
on the prepared throne, flanked on either side by the symbols of
the Evangelists placed respectively above and below these hori-
zontally stratified clouds. Contiguous bands of red and yellow
indicate the dawn; at least that was probably the suggestion
which they were intended to convey. The selection of the colours
is in itself significant.
C. Woven and embroidered Stuffs. Among the most
surprising discoveries resulting from careful excavation are the
silk and woollen stuffs, to the preservation of which the sand of
Egypt has conduced even better than the care lavished upon
them in museums. They agree in every respect with those
Western examples with which we have long been familiar
through examples in mediaeval reliquary-shrines. The monu-
ments recently discovered (in Central Asia) in Chinese Turkestan,
especially in Khotan and as far as Tunhuan^, which have been
similarly preserved by absolutely dry conditions, confirm us in
the belief that we have to deal here with an industry and form of
decoration not peculiar to Egypt, ^ but having a perfectly homo-
geneous distribution round about the central area of Iran. These
textiles rarely deal with Christian subjects, but for the most part
distinctively retain that spirit which I have already tried to
describe in the section on Mazdean costume. The large majority
of the silk stuffs depict with mirror-like fidelity Persian horsemen
engaged in the chase, or, as a substitute, a formal tree of palmettes ;
these are invariably disposed in a system of interlacing circles.
The woollen textiles are geometrically divided by interlaced
bands into panels filled with figures of animals and birds in the
varying combinations usual to Hvarenah symbols. It has often
been conjectured that these textiles may have suggested the
predominant style of architectural decoration as well in Lombardy
as in Scandinavia. But in point of fact — quite apart from what
^ At this point brief reference may be to distort the facts for students (of.
made to Otto v. Falke's Kunstgeschichte Amtliche Berichte der preussischen Kunst-
der Seidenweberei, a work of great value sammlungen, xl [1919], pp. 143 ff.)- A
through the amount of comparative work of my Institute (Dimand, Die
matenal which it brings together, and Verzierung der koptischen Wollivirkereien,
its weahh of illustration. I have stated Stromungen des Weltverkehres im Kreise
my position with regard to it in Altai- der Mittelmeerkunst) might help to
Iran (see Index under Falke). The restore some sort of order amid the
author had not a sufficiently wide range prevalent confusion, but present condi-
to be a sound judge in questions of tions make an early publication im-
development, and he makes untenable possible,
assertions as to origins, which continue
53 Armenia, Ketcharus monastery ; South door of Church of S. Gregory
See p. 148.
54 Dashlut, Egypt ; door of Court of Mosque ; the mounted saint,
half destroyed. See p. 175.
55 Vienna, in private possession ; woollen fabric from Egypt. See p. 153.
56 Urnas, wooden church ; door with interlaced beasts ; after A. Haupt.
See pp. 153, 215.
THE ANTI-REPRESENTATIONAL MOVEMENT 153
was indigenous in the North — this obvious affinity is adequately
explained by the connexion of Iran with the extreme North as
well as with Egypt and upper Italy ; and in this matter the agency
of workers in mosaic and stucco was no less effective than the
trade in the textiles themselves. I have given an illustration of
a woollen fabric from Egypt (Fig. 55, opposite), showing the
Iranian vase with scrolls in the centre, flanked by a double row
of formal trees, animal figures, and medallions.
IV. The European North. The path pursued here was
essentially similar to that of the East Aryans. True, it is hardly
possible for us now to imagine this area as having ever been
devoid of representational art ; and yet at the end of the first
millennium it was still devoted to a purely decorative style.
Let us consider the Scandinavian wooden churches such as
Urn as. The gable decoration reveals an almost hypersensitive
delight in expressive line : the Greeks filled this triangular
space with human figures. Or again, consider the oldest surviving
woodwork on the exterior of the same church (Fig. 56, opposite),
which is among the most delicate memorials bequeathed to us
by the soul of the North. In such a spiritual revelation the
human figure would have a crude effect. Salin has solved the
riddle of certain ancient Germanic bronze ornaments which betray
a similar delight in the play of line, the inspiration coming here
too from animal forms. We do indeed find on the portals of
wooden churches isolated examples of scenes from the myth of
Siegfried depicted in combination with scroll patterns. But
instances of this kind are rare and late ; moreover, it is necessary
to distinguish between realistic and imaginative figures. Imagina-
tion will be found to be the basis of every work of art both in
Iran and the North, as well as in all artistic movements connected
with them and uncontaminated by Southern influence. The
human figure may be sporadically introduced in conjunction
with animal and plant motives. But, like these, it is never treated
objectively with continual reference to a natural form ; it is
simply an intelligible sign written in a readable hand.
The point to remember is that the North, like the East,
employed in the decoration of its buildings a system which ex-
cluded the human figure. The style of the finds made in the
Oseberg ship, finds of unprecedented richness, differs from that
of the prehistoric bronze ornaments and that introduced by the
Goths and Lombards into Italy ; these Oseberg objects are
154 NON-REPRESENTATIONAL CHURCH ART
decorated not with interlaced patterns purely geometric in char-
acter like the Armenian, but with that luxuriant riot of animal
forms, with which the important publications of Sophus Miiller
and Salin have made us acquainted. The East was as unable
to follow the North on this path as in later mediaeval times
it was unable to share the vigorous contribution made by the
Northern spirit to the decoration of Gothic cathedrals. In these
cathedrals the dominant quality is the feeling for organic structure
possibly carried over from earlier architecture in wood ; it makes
itself felt in the smallest details ; the ever urging impulse lends
every point and edge the luxuriance of plant growth, just as, long
before, the same ferment in the Northern nature had covered
every inch of space with expressive animal forms. In Armenia,
which took half a step in the direction of ' Gothic ', we find no
trace of this more recent movement in decoration.
I shall refrain here from pointing out evolutionary affinities
in the decoration (as I attempted to do in the architecture) of
the * Romanesque ' churches in the West, which I call oriental.
Such an attempt would carry me beyond the scope of this volume.
VII
The Triumph of Representational Art.
Hellenism, Semitism, Mazdaism
FORMATIVE art employs visible signs as intermediaries
between the direct effect of mass, space, light, and colour,
and particular meaning and purpose. Religious art, more
especially, has a predilection for the most familiar of all symbols,
the human figure, which seems, like the actor, to bridge the gap
between life and art. At the same time we have observed that
religious areas of considerable extent renounce this mediation ;
they rely solely upon structural form, and adorn their buildings
with decorative designs which are often of a symbolical character.
In Christian art the practice of representation appears to have
been introduced as a result of contact with Hellenism and
Buddhism. To depict Christ as a human figure was not such an
obvious idea to Jews, Armenians, Arabs, or Iranians, as it was
either to men inheriting the traditions of the great Semitic
empires, or to Greeks and Romans. As late as the tenth century
the Armenian historian Thomas Artsruni, in referring to a picture
representing Christ in a manner quite familiar to us, thought it
necessary expressly to state that it was the Saviour thus depicted
in human form. It is significant that we in the North require
to have our attention deliberately drawn to the fact, before we
can see anything surprising in the use of the human figure, in
which our religious art of the present day finds almost its sole
expression.
We distinguish between realistic and conventional representa-
tions. Down to the thirteenth century Christian, as opposed to
ancient Semitic and Graeco-Roman art, practised only conven-
tional representation. In its infancy the representational art of
the Christians in West-Aryan regions was admittedly much
influenced by good Hellenistic models. Hence originated a
classical-Christian style, of which the Anatolian type of Christ
is an example ; in this type the head might be the work of a
156 TRIUMPH OF REPRESENTATIONAL ART
Praxiteles. The figure of the Teacher on the BerHn sarcophagus
reproduces the antique type of the orator; the familiar marble
statue of Sophocles reminds us what the original was like. In
the flourishing city-life of Northern Mesopotamia, in Edessa
and Nisibis, more remote from the heart and pulse of Hellenism,
we find a very diff^erent form of expression. There the national
type of Christ early assumed Aramaean features ; the straight
hair was parted in the middle, and the beard was long. The
influence of the powerful Semitic states led to this kind of realism,
which was accompanied by an historical conception of art. The
reduction of naturalism in Christian-Semitic representations was
principally the work of the third great source of artistic influence,
the Mazdean.
It has been asserted that a diminishing fidelity to nature and
its wholesale reduction to geometrical forms, a process already
observable in late-Roman art, is neither more nor less than a
confession of artistic bankruptcy. This is a mistake. Roman
powers of observation were slow to weaken ; but Eastern Iran
had introduced a new taste, which completely reversed the ancient
trend of art. The tendency to ' geometrize ', in the East- Aryan
manner, might almost be described as convalescence in art.
Art began to develop afresh from within, and to use new imagina-
tive forms of expression ; these were applied exclusively to the
surfaces of buildings and designed to impress the spectator from
a distance, differing in this from figure sculpture, which is
calculated for a closer view. For this reason alone, if for no other,
it is necessary for any one approaching the subject of Christian
art to begin with the Churches and not with the Catacombs and
the Sarcophagi.
Eastern Iran had always secured its effects chiefly by the
combination of mass, space, light, and colour, at the same time
supplementing these with a non-representational form of art.
The Jews among whom Christianity arose themselves inclined
to this form of art less on account of their Semitic origin than
because they were pastoral nomads at the time when their own
religion was founded, and they retained the mode of expression
inspired by a desert life. The Semites of the empires holding
the great river-valleys became uncompromising exponents of
representational art ; in fact it was probably among them that the
tendency towards portraiture originated. This was a natural con-
sequence of their compact organization and their approximation of
HELLENISM, SEMITISM, MAZDAISM 157
the ideas of God and king ; they depitt not only their rulers but
even God with individual features ; while in scenes depicting life,
they show a tendency towards naked realism. The origin of
historical scenes must also in all probability be ascribed to them.
They played an important part in the growth of Hellenistic,
Buddhist, and Christian art by bringing about the transition from
symbolism to representation, whether of the king's own life
or of biblical events. It is, therefore, hard to understand how
any one can assert of Hellenistic culture that it was oriental only
in its political and monarchic basis while artistically it remained
Greek. Not only is it misleading to make ' representation ' stand
for art as a whole, but it is a mistake to regard as inherently Greek
an art based on monarchical conceptions and only temporarily
clothed in a Hellenistic garb. Christianity re-introduced the
ancient oriental flat treatment with its ' primitive ' handling of
space, mass, light, and colour ; and in particular it revived the
continuous arrangement of subjects in rows, the object of which
was to represent the hero (whether ruler or religious Founder)
with overpowering effect as the perpetual centre in a long and
varied sequence of illustrations.
In this respect art must have been decisively affected by
that focus of ecclesiastical influence which must be placed some-
where midway between the East Mediterranean littoral and
India, because Buddhism passes from symbolism to historical
representation in the same way as Christianity. This transition
was consummated in the period between the purely Indian art
of the stupas of Barahat and Sanchi and that of the Gandhara
sculptures, the style of which stamps their origin as Hellenistic
beyond question. But they are so only in outward appearance ;
their underlying spirit remains no less Semitic than that of the
Christian art of the Mediterranean in its transitional period as
seen during the fifth century. The experiences of the last few
years incline me to assume that it was masters from northern
Mesopotamia who introduced the practice of representing a series
of events in the life of a hero alike into late classical, Buddhist, and
Christian art. Whether we consider the spiral reliefs of triumphal
columns at Rome and Constantinople, illuminated rolls of papy-
rus or parchment, or the sculptures of the Indo- Afghan frontier ;
whether the subject be emperor, Christ, or Buddha, we invariably
find the same blending of Semitic ideas with a Greek or Indian
style. I shall first examine the features which distinguish
158 TRIUMPH OF REPRESENTATIONAL ART
symbolic from realistic representations, in so far as their traces
are still discernible in the East ; I shall then attempt to locate
the point at which the transition was effected, so far as the present
state of our knowledge permits.
I. Symbolic representation of the West- Aryans
Wherever the Aryans came into contact with the art of the
South and adopted the representational style without losing
their spiritual predominance, we find the human figure treated
less as the embodiment than as the symbol of divinity. The
greatest reserve in this matter was practised by the I ndo- Aryans,
who originally avoided representing God or the Founder in human
form, but employed suggestive symbols such as the wheel, the
Bodhi-tree, or the stupa. Such symbols are indeed ubiquitous ; the
peculiarity of Indian art consists not so much in the symbols
themselves as in the attitudes of the groups of human figures
surrounding them, which indirectly suggest the divine presence.
Max Klinger has recently adopted a similar method, though in
a somewhat different sense, in his sketches illustrating the Sermon
on the Mount ; here the meaning is not directly expressed by
the figure of Christ in the act of preaching, but suggested by
the contrast between what went before and what followed. The
Greek generally agrees with the Semite in getting into direct
contact with his subject, only differing in his conception of God ;
the Semite views Him as the embodiment of power, the Greek
as the perfection of manhood. A change indeed occurred in
Hellenistic times. Nevertheless, in those Greek regions which
were not actually imperial residences, such as Alexandria and,
later, Asia Minor, the ancient beauty of the human figure in
some degree survived and was transmitted to Christian art.
The persistence of this Hellenistic current of symbolic representa-
tion may be seen in the Joshua Roll, the Psalter 139 in Paris,
the reliefs of the Sidamara Sarcophagus at Berlin, depicting
Christ as an orator, the Lipsanothek in Brescia, and the statuettes
of the Good Shepherd. An example of the Hellenistic treatment
of a Mazdean idea is, in my opinion, also to be seen in the ivory
carvings of Yima, the good shepherd^ {supra, p. 122) ; and in
this connexion we should also remember the Mithraic reliefs.
1 For further information on this entitled Orpheus, und verwandte iranische
school of art the reader is referred to the Bilder.
appendix of a short work by O. Kern
57 Island of Achthamar, Lake Van, cruciform church ; West side.
See pp. 150, 159.
HELLENISM, SEMITISM, MAZDAISM 159
This Hellenistic character is dominant in Early Christian
sepulchral art at Rome, permeated though it is with the ideas of
intercession and redemption. Both in the Catacombs and on the
Sarcophagi human figures alone convey the symbolic meaning,
and, on the latter, they are crowded together in a most inartistic
manner. Despite the obviously Greek treatment of forms and
drapery, the influence of another school of art is apparent,
a school which knows nothing of individual values, and seems
in a fair way to treat human figures as no more than light pattern
diff^used on a black ground. Representation is in fact retained
only by the imperative demand for a significance in which the
mere object and the endless mechanical repetition of symbols
count for more than content.
That this art was not unconnected with the East in its origin
and growth^ is proved by a monument hitherto ignored by
Christian archaeologists, but of exceptional importance, in spite
of its late date. I refer to the exterior decoration of the" Armenian
cruciform church of Achthamar on Lake Van, to which allusion
has already been made. This building, erected between the years
A. D. 904 and A. D. 938 by King Gagik of Waspurakan, has the lower
part of its exterior entirely covered with reliefs, which are probably
not to be connected with the Iranian animal frieze discussed
above. On the West side (Fig. 57, opposite) the king himself
is represented standing before Christ and holding a model of the
church. OntheEastside(Fig.39,facingp. 122) are the evangelists (?)
and saints. But the reliefs which concern us here are those on
the North and South sides. They too breathe the same spirit as
the animal friezes ; interspersed between the biblical scenes are
whole rows of Hvarenah animals one above the other, but in
the present place I shall discuss the figure scenes alone. They
are inspired by the same group of ideas as the paintings in the
Roman catacombs, the ideas conveyed in the Easter prayers, or the
prayers for the dead, and the instances of redemption enumerated
therein, more especially those from ^he^ Old Testament. At the
AVest end of the South side (Fig. 33 , facingp. 1 14) tEestory of Jonah is
depicted at considerable length, but not in the classical Christian
style. In the scene of Jonah addressing the Ninevites, the king is
seen full face, seated with crossed legs ; but the arrangement of
* Jewish prayers for the dead and the Alexandria with its large Jewish popula-
thoughts proper to the Easter festival are tion is assumed as the point of departure
often held to have influenced this art.
i6o TRIUMPH OF REPRESENTATIONAL ART
the figures is still the old familiar one of the Catacombs and
Sarcophagi. The same applies to the scene of David and Goliath,
in which Saul stands on one side in caftan and turban. Scenes
from the story of Samson lead round the Church to two pictures
of Adam and Eve and the Fall. The Western corner of the North
wall opposite the Jonah pictures completes the circuit with
figures of the three children in the fiery furnace and Daniel in
the lions' den. Circular panels containing figures of prophets,
saints, some of them contemporary, and of Christ and the Virgin
enthroned side by side, are inserted between these biblical scenes.
The richness of these sculptures is quite un- Armenian in character.
Achthamar is evidently an example of the blending in a frontier
region of Mazdean Hvarenah ornament with that style of
sepulchral art which reached Rome from the Syro-Egyptian
region in the first centuries of our era. When we remember that
the national form of Armenian church was emphatically of the
memorial type, and stood in the midst of a large courtyard, we
can understand why the exterior of the cruciform church of
Achthamar gives expression to the hope of redemption. Here
we may assume a new artistic province, for which few proofs
are elsewhere available. My theory is that it was the custom to
decorate the exterior surfaces of churches built of unburned
brick with flat reliefs in stucco, as well as with painting. Examples
of both these styles exist in the geographical area covered by the
Orthodox Church. They are not actually built of unburned
brick faced with stucco, but, like Achthamar, imitations in stone.
I refer to the churches of Vladimir and Yuriyeff Polskiy in Central
Russia. Their outer walls, dating in the latter case from the
years a. d. 1230 to a. d. 1234, are covered with a profusion of
flat reliefs, disposed in zones under blind arcades, and illustrating,
though in a much less rigorous sequence, the same group of
ideas expressed at Achthamar. The Church of S. Sophia in
Trebizond afi"ords another example. Here on the south side is
a band of reliefs with scenes from Genesis. Examples of the second
or painted type of decoration may still be seen in the Rumanian
monastic churches of the Bukovina, dating from the sixteenth
and seventeenth centuries (Fig. 58, opposite). In these buildings
we invariably find the Last Judgement on the West wall ; while
the North and South walls are decorated with a large series of
pictures illustrating the Tree of Jesse, the Akathistic Hymn, the
Ladder of John Klimakos, scenes from Genesis, and other
o
\0
<D
I
3
O
CO
o
3
C
o
c
'>
o
-a!
3
ca
C
o
o
>
00
k
HELLENISM, SEMITISM, MAZDAISM i6i
subjects. On the trefoiled East end we find figures of saints in
superimposed zones, in the uppermost a band of angeHc choirs ;
all these figures are grouped about the Trinity.
We have here apparently the monastic version of an old tradi-
tion. This type of art has no connexion with the West, but it occurs
in the monasteries of the South, for example at Mount Athos,
in the porches and refectories. It brings before the eyes of the
monk that which should form the subject of the day's reflection.
The only affinity between these and Early Christian pictures
lies in the manner of their application ; we may recall their
absence in the account of Neon's refectory at Ravenna. Un-
happily these things have never been brought into their true
relations. My only reason for introducing these examples here
is to illustrate their connexion with the monumental art of
churches and monasteries and that fundamental love of symbolism
which marks the Early Christian period. Nevertheless they are
characterized not so much by the ideas of intercession and
redemption as by a didactic spirit, for the origin of which I am
inclined to look in another direction.
II. Semitic Realism
The idea of God among the Semites living under the
empires in the great river valleys appears to have been directly
inspired by the ruler's lust after power. The divine figure is
furnished with all the attributes of the despot ; he receives the
marks of homage for which the earthly monarch longed. In
Christian times the true carriers of Semitic culture were the
Aramaeans of Mesopotamia, who played a part similar to that
of the Arabs and Nabataeans in Islam. These Semites were
indeed subjected successively to Roman and Persian rule, but
their old love of the royal title, conferred in ancient times upon
every little princeling, now found its expression in the field
of religion. It was no doubt in this cultural area that the great
type of Christ Pantokrator developed from the Hellenistic figure
of the judge. This manner of representing the Almighty seemed
so natural to Semitic Hellenism that theologians found no
difficulty in applying it to the person of the Founder. This
figure gained general acceptance, since it avoided all controversy
as to his human or divine nature. Compare with it the type of
Buddha seated cross-legged and sunk in inward meditation.
2451 ' M
i62 TRIUMPH OF REPRESENTATIONAL ART
In India Buddha is never depicted with the ruler's attributes ;
he is never enthroned, nor does he wear the severe aspect of the
judge holding the Law in his hand. How different the type of
Christ.
The type prevalent even among ourselves to-day, that with
beard and long smooth hair, is a Semitic creation. The Founder
bears the physical characters of his race, like the first portrait
heads of Peter and Paul, which in the same way have never
changed. The two Hellenistic unbearded types of Christ, the
long-haired of Asia Minor and the short-haired of Alexandria,
have been completely superseded in the West. The Aryans, in
fact, abandoned the Greek figure of the Founder in favour of the
Semitic type of God personified as the Ruler. The true origina-
tors of this change were the Aramaeans, and they were closely
followed by the Byzantine Court. As far as origin is concerned,
this remarkable sequence of development is proved by the fact
that the Pantokrator, who embodies the popular idea of the Judge
of the World, assumes this Syrian type ; as far as distribution
is concerned, by the fact that Christ is early surrounded with
the emblems of royalty.
The Aramaeans were not only settled in northern Syria ;
they migrated to Mesopotamia also ; and it was precisely in
ancient Assyria, bordering on Armenia, that the moral and
intellectual influence of this people was at its strongest. Their
significance for Christianity is due partly to this central position
between the Mediterranean and Persia, but even more to the fact
that, when the Persians overran the Near East, Aramaic became
the official and commercial language of the western half of their
empire. Thus cultural unity was established throughout the
southern territories of Persia, extending from the Indian frontier
over the Sassanian middle region to hellenized Asia Minor ;
we understand why Edessa and Nisibis reached the paramount
position which made their territory the very heart of Christianity
in the East, why it actually became the starting-point of western
monastic activity, through Cassiodorus' foundation of his monas-
teries in Southern Italy.
Christian Semitic art derives its spirit, indeed, from Semitic
tradition, but its outward form in great part from Hellenism.
Its root-idea is to instruct, and its use of the human figure is
classical only in so far as the ancient oriental world allies
itself with the Hellenistic East. This current appeared in Chris-
HELLENISM, SEMITISM, MAZDAISM 163
tian art at a later date than the non-representational art of the
East- Aryans, and combined with the latter to form that whole
which we call ' mediaeval ' art. In this mediaeval art, an art far
from nature, the symbolism of the West- Aryans of the^outh only
reappears sporadically in genuine Renaissance movements, which
occurred in Byzantium as well as in the West. I shall now examine
first the meaning and subsequently the outward form of this
Semitic-Hellenistic province of art.
A. The Didactic Element. The first change in Christian
Hellenistic art, the true home of which in the first three centuries
lay in Western Asia Minor rather than Alexandria, took place in
Antioch, the great port of Asiatic commerce, and thence was
transmitted to the Mediterranean. Symbolism was displaced
by a didactic tendency, the purpose of which was to teach the
faithful by the help of graphic art how the teaching of Christ
and the prophets was handed down through the apostles and
evangelists to the Church. This purpose is clearly expressed in
fifth-century mosaics such as those of the Mausoleum of Galla
Placidia and of the orthodox Baptistery at Ravenna. In the
former the cross is shown at the crown of the dome, and the
illustration is completed with the figure of the Good Shepherd
bearing the cross, and the burning of the writings condemned
by the Church ; in the latter, the Baptism of Christ is appro-
priately represented in the dome ; this is succeeded by the Church
and the prophets, while the scheme was formerly completed by
the four subjects in the lower corner-niches ; the division into
compartments by formal trees may be compared with that in
the Church of S. George in Salonika and that in the Church of
the Nativity in Bethlehem. The conception of this purely
ecclesiastical system is most characteristically expressed in the
scenes representing the Giving of the Law and the Keys, as they
appear both in mosaics at Rome and in the sculptures of the
Sarcophagi at Rome and Ravenna.
This influence from Antioch must be distinguished from
another, to which the reader may best be introduced through
that letter of Nilus, the first part of which has already been
quoted on p. 144. In reference to the Iranian non-representa-
tional style, which we discussed above (pp. iii ff.), he continues :
* To this communication I can only reply that no one but a babe
* or a suckling would wish to pervert the eyes of the faithful
* with such trivialities as those which you describe. It beseems
M 2
i64 TRIUMPH OF REPRESENTATIONAL ART
a manful and firm mind to place in the apse, at the East end of
the house of God, nothing but the cross alone ; for by the single
cross of salvation mankind is saved and hope universally
Eroclaimed to the hopeless ; but it is not unfitting to decorate
oth sides of the holy temple by the hand of the finest painters,
with scenes from the Old and New Testaments, that the man
who is ignorant of writing and unable to read the holy scriptures,
may gaze upon the painting and gain knowledge of the fathers
of all virtue, who served the true God ; and that he may thus
be roused to emulate them in the great and celebrated deeds
of heroism, through which they won heaven in exchange for
earth, setting the things which are unseen above those which the
eye perceives. But in the nave with its various chapels, each
chapel should display the sublime cross alone. Anything beyond
this should in my opinion be excluded.*
It will be observed that the letter consists of two sections :
the first is a restatement of the question asked by the eparch
Olympiodorus. The second is a violent attack on the proposal,
quite ingenuously made, for Iranian forms of ornament, to which
is appended a reasoned counter-proposal for a decoration after
the Semitic manner. Nilus is thus a living witness of the
transition from an individual and freely developed religious art
to an art under ecclesiastical control. The time at which he
wrote seems to have been an ideal moment for some such
change. Christianity, which had begun by making its way in
small communities, and had then been raised to the position of
a state religion, was now devoting all its energies to the struggle
for power and the complete subordination of the peoples to the
authority of Church and Empire. The artist was degraded by
them to the position of a mere instrument. He was no longer
permitted to decorate freely, but forced to depict what the Church,
and subsequently the Court or the combination of Court and
Church, prescribed. Thus the character of religious art was
changed ; the talent of the artist was exploited by authority ;
his personal freedom was destroyed.
The most important demand formulated by Nilus, which
finds its parallel in similar utterances by the Fathers of the
Western Church, was for a form of mural decoration in churches
which should serve to teach the illiterate believer. The struggles
and sufferings of the martyrs had been depicted in apses before
A. D. 400 ; now the Church demanded that the opposite walls of
HELLENISM, SEMITISM, MAZDAISM 165
the nave should be decorated with parallel scenes from the Old
and New Testaments. The Semitic historical sense was beginning
to triumph over East-Aryan non-representational feeling and
West- Aryan love of symbolism by means of the human figure.
For the purpose of determining the date, place, and culture in
which this tendency originated, it is worth noting, as has already
been pointed out, that it is also to be traced in the Buddhist art
of Gandhara. In studying this didactic manner we may best
begin with ivory carvings, the very material of which points
towards the East.
The chief work of this class, the chair of Maximianus, has
already been mentioned (p. 151) in the section dealing with the
earliest phase of Christian art in the East, as an illustration of
rich decoration with scrolls and animals. But even more im-
portant, perhaps, than these are the figure-subjects which fill
its panels. Like Achthamar, but earlier by four centuries, it
illustrates in a most instructive way the fusion of Iranian orna-
ment not so much with the purely symbolic as with the Aramaean
didactic style. On the front of the chair, enclosed between
borders of scrollwork containing animals, are figures of those
who spread the faith, grouped on either side of John the Baptist,
last of the prophets, represented bearing the lamb. In their
treatment these figures resemble those of the sarcophagi from
Asia Minor, which stand between the columns of arcades with
scalloped canopies, and reveal the dearth of contemporary models ;
for though one sarcophagus belongs to the Christian period,
all betray imitation of Greek sculpture dating from the fourth
century before Christ. A related ivory panel, now in the British
Museum, with the figure of an archangel and a Greek inscription,
points more definitely to Antioch as the source of inspiration ;
the figure is placed in front of the proscenium in a manner
resembling the priest before the altar-screen at a later date
(cf. the Pola casket). We are carried a stage further by the panels
on the back of the chair, carved with low reliefs illustrating
the life of the youthful Christ. Both in their limitations and their
freedom these carvings closely resemble the five-panelled diptychs.
These very diptychs show the same remarkable arrangement
of a single central figure surrounded by scenes from Christ's
childhood on the upper, lower, and lateral panels. They resemble
the front of the Ravenna chair in almost every point except that
the scrollwork is replaced in the upper panel by a figure of Christ,
i66 TRIUMPH OF REPRESENTATIONAL ART
or by a triumphal cross borne by hovering angels, and in the
lower by an Adoration. They clearly belong to the same school
of art. We can no more ascribe their origin to Ravenna than we
can regard the two in the Ravenna library as Coptic. I agreed
with these attributions until my eyes were opened by Mshatta
and Amida ; I see yet more clearly now that I have been able to
study Altai-Iran and Armenia. The ivory of which the diptychs
are composed suggests a connexion with India ; moreover the
arrangement of the scenes in vertical and horizontal bands grouped
about a central figure finds such a marked analogy in the Gand-
hara reliefs that it can hardly be explained by independent
invention consequent on the inevitable likeness between two such
kindred subjects as the lives of Christ and Buddha. Hitherto
I had always regarded Jerusalem as the centre of distribution
for this type of sculpture, without perceiving its affinities with
India ; I now recognize that it was probably a posthumous
influence from the Graeco-Semitic centre of Edessa-Nisibis,
which here affected Christian art as, at an earlier period, it had
affected that of Northern Buddhism. It would take me too long
in the present place to demonstrate the similarity, hardly yet
noticed, between the Gandhara sculptures and the didactic
school of Mesopotamia and, later, of Byzantium and Europe.
The relationship of their subject-matter has already been pointed
out by Metzger-Milloue.^ My immediate purpose is to give
a fresh and somewhat fuller account of the iconographic cycle ;
and in doing so I shall treat the types of sculpture and their
arrangement separately.
B. Aramaean Cycle. Since no example of an Early Christian,
or, in particular, of a Nestorian church with its original paintings
has yet been discovered in Mesopotamia, we have to rely on the
evidence of rock caverns, though these are only of secondary
value. For paintings we shall begin with the Catacombs of
Palmyra and Edessa ; for sculpture with those of Dara. In the
latter place the entrance to the rock-cut tombs is decorated with
an arch enclosed by a n-shaped moulding, a form of ornament
the Iranian origin of which is repeatedly impressed upon
our minds. In one instance (Fig. 59, opposite) the spandrels
between the arch and the moulding are filled with sculptures in
relief, among which is an undoubted representation of Christ in
1 Les quatre dvangiles. Matdriaux pour servir a Vhistoire des origines orientales du
christianisme, 1906.
59 Dara, rock cave ; exterior. See p. i66.
---.>..
60 Berlin, Kaiser Friedrich Museum ; mounted Christ from Upper Egypt.
See pp. 174. 175.
iP f^;'^#
.p *>
''.i*..
61,62 Florence, Laurentian Library ; MS. of Rabula, a.d. 586 ; Eusebian
Canons ; photo., G. Millet. See p. 167.
HELLENISM, SEMITISM, MAZDAISM 167
Limbus. He is depicted striding rapidly towards small naked
figures of Adam and Eve, which appear above a mound of skulls.
In the corner above and behind Christ is the hand of the Almighty.
This is probably one of the earliest surviving representations of
this subject, which later became so popular in Byzantine art, and
was frequently associated with the scene of the Last Judgement.^
The subject in the opposite corner is no longer recognizable,
but its chief feature is a large cypress. The arch beneath is orna-
mented in the regular Graeco-Persian style of the Christian
period in Northern Mesopotamia. May not these Mesopotamian
sculptures have been the prototypes of the Achthamar cycle ?
Another important monument from this region is the
Laurentian manuscript in the Laurentian Library at Florence,
written by Rabula in A. D. 586, and originally in the monastery
of Zagba. The Eusebian Canons of the Evangeliary are famous
(Figs, 61, 62, opposite). The arches are enriched with geometric
ornament ; flanking them at the top are figures of the prophets ;
below these are scenes from the New Testament arranged in series,
and at the base Iranian animals and plants, including vines and
pomegranates. It is the richest decoration of its kind surviving
from so early a period ; and, taken in conjunction with the ampullae
from Jerusalem at Monza, in which the treatment of the figure-
subjects is similar, it is very important evidence for the derivation
of later Byzantine iconographic motives from the religious centre
of Northern Mesopotamia : it was there that this special type of
art was devised for the instruction of those who could not read.
Other Syrian manuscripts of the same type exist, for instance, in
Paris ; there are also Armenian and Greek manuscripts which
follow the same lines. But the arrangement of the figures about
the arcades is not retained. They are generally placed after the
Canons at the beginning of the separate gospels, and often take
the form of whole-page miniatures. To any one familiar with
the mass of illustrations in the Oktateuch, the derivation of which
is known, the extraordinary profusion of miniatures in the
Gospels with illustration in zones or friezes will hardly come
as a surprise. On some pages two or three superimposed zones
interrupt the text. The important points to notice are that the
composition of individual scenes is always the same, and that the
scenes illustrating the childhood of Christ, as in the diptychs
and on the episcopal chair, are based upon the apocryphal
^ e. g. in the Torcello Mosaic.
i68 TRIUMPH OF REPRESENTATIONAL ART
gospels. Most striking of all is the arrangement of the Nativity
scene, including the figures of ox and ass, Joseph in an attitude
of meditation, and frequently the bathing of the Infant. It is
impossible for me here to enter into the details of even a few
of these innumerable illustrations. For points of detail, published
studies in iconography should be consulted.
I am thus driven to assume the existence in Northern
Mesopotamia of a school of representational art, in which early
streams of Eastern-Semitic and Greek culture met and inter-
mingled. Its influence extended first to India and later to the
Mediterranean, and is readily distinguishable from the East-
Persian non-representational type of art which I have already
discussed. Its style is that of a continuous narrative, in which
the central figure is consistently thrown into relief, whether in
the character of Christ, the Virgin or Joseph, saint or martyr in
Old or New Testament.
This was the centre which initiated the parallel arrangement
of Old and New Testament scenes on opposite walls demanded by
Nilus. Its place of origin was possibly the same as that of
the Eusebian Canons, in which the parallel passages in the gospel
are arranged opposite each other. In the surviving monuments
of Early Christian art we cannot indeed point to any example
of such an arrangement of parallels from the Old and the New
Testaments ; but the story of Joseph on the sides of the episcopal
chair is depicted with as much detail as that of Christ and the
Virgin.
The proper place in churches for this arrangement of parallel
subjects was on the walls of the nave, a fact attested by others
besides Nilus. No example has been preserved, unless we
count the Church of S. Maria Maggiore, where scenes from
the Old Testament are placed in the nave, scenes from the New
Testament on the triumphal arch. It is noteworthy that the
latter breathe the same spirit of the apocryphal gospels, from
which the reliefs of the ivory diptychs and the chair of Maxi-
mianus derive their peculiar charm. If my interpretation is
the right one, and the oldest and most important of the five-
panelled diptychs in the Louvre represents Constantine in the
centre as the champion of the Faith with the Caesars on the
lateral panels, the conception of the rider triumphing over evil
indicates an eastern origin and contact with Mazdean ideas :
I shall revert to this matter below. On the other hand, the treat-
HELLENISM, SEMITISM, MAZDAISM 169
ment of the reliefs on the rest of these diptychs is distinctively
Semitic. In the mosaics of S. Maria Maggiore yet another
influence is apparent, particularly in the figures of Christ and the
Virgin, which are invariably accompanied by angels and a splen-
dour of costume obviously derived from the pageantry of court
life.
C. Sequence of Didactic Subjects. I have pointed out how
necessary it is to distinguish between type and composition.
The latter is as variable in church decoration as in the manu-
scripts; and I believe that the Early Christian subjects of the
Catacombs and the Sarcophagi, where all centres on the idea of
redemption, were modified under Aramaean influence. I base
this conclusion upon a study of the reliefs and paintings
on the exterior of the churches discussed in connexion with
Achthamar. They do not introduce any new set of ideas, but
merely serve to transmit the old symbolism of the Catacombs
and Sarcophagi to later times. Second in point of time comes
that province of art which I discussed in connexion with S. Nilus ;
it is associated with the architectural form of the basilica, in which
not only the apse but also the walls of the nave above the arcades
were decorated with scenes from the Old and New Testaments.
In the arrangement of subjects, a leading part was played by the
theological school of Edessa and Nisibis. Whether the introduc-
tion of the concordance between the two Testaments may be
attributed to its influence is a question on which I must reserve
judgement. But there is no doubt that it created the cycles
familiar to us from innumerable domed churches, and from the
instructions in the * Painter's Manual ' of Mount Athos. It was
first designed for the single-domed church, and probably
originated, like the domed basilica, on the Armenian frontier.
This iconographical scheme received general recognition as an
integral part of every domed church, but Jerusalem was the
starting-point of the individual figure types, which reached
Western Europe through Byzantine channels and Eastern Asia
through the missionary activity of the Nestorians, thus acquiring
world-wide significance. The essential point remains the
association of Aramaean iconography with domed architecture,
and its achievement in these churches of a decorative system so
logical and perfect that in its quality of immovable persistence it
stands without a rival. The first traces of the system are charac-
teristically found in Ephraim the Syrian's picture of the Last
170 TRIUMPH OF REPRESENTATIONAL ART
Judgement. We do not, indeed, meet with a full union of this
iconography with domical architecture before the Nea, the
palace church, built by Basil I in a. d. 88c ; this emperor was the
first of the Armenian dynasty to ascend the throne of Byzantium,
and may well have employed an Armenian architect in the con-
struction of this church, which is of the characteristic cruciform
domed type. Its decoration may likewise have reached Byzan-
tium from a foreign source. It is well known that there was
a figure of Christ Pantokrator in the apex of the dome and of
the virgin as Oram in the apse, accompanied by rows of apostles,
martyrs, prophets, and patriarchs together with angelic choirs.
The much later * Painter's Manual ' from Mount Athos prescribes
in detail the exact position of all these figures, adding instructions
for five superimposed zones of paintings for connexion with these
subjects. I shall not pursue this subject further, only remarking
that the ' Painter's Manual ' also distinguishes the types of
figures from their arrangement.
This rapid survey must suffice for the didactic iconography
of Edessa-Nisibis. It triumphed in Constantinople and in the
whole of the Mediterranean area, as well as in Armenia and the
Orthodox Greek Church. In the latter it survives unchanged
to the present day as the ideal both of type and arrangement ;
but in the West it was superseded as early as the fourteenth
century. A counter-movement started in the North and thence
spread into Italy. It is interesting to notice how Duccio, Cimabue,
and the elder Pisano were forced into the background by their
own pupils under Northern influence.
Reviewing the course which I have followed since the publica-
tion of Cimabue und Rom in 1888, I am struck by the need for
a work dealing with the penetration of the South by the North
of Europe and the subsequent rise of the so-called Renaissance.
Burckhardt was mistaken in claiming that the artistic emancipation
of the individual took place in the South : on the contrary, it is
Northern. Even if we maintain that the Renaissance was a purely
Southern movement, in Italy it only matured by slow degrees,
at first under the artists employed by the Hohenstaufen, then
through the adoption of classical ornament in architecture, and
finally by the imitation of statues and the study of the orders.
Between Cimabue and Giotto, the elder and the younger Pisano,
intruded the North. It began with wooden architecture and pure
decoration. The Teutonic beast-ornament, discussed above, can
HELLENISM, SEMITISM, MAZDAISM 171
in no sense be regarded as representational art. The latter grew
independently out of the organic impulse of Northern church-
art (Gothic), first in opposition to plant motives, in the form of
animal finials and gargoyles ; of this development Achthamar
may be regarded as an enigmatical forerunner. The human
figure also appears as an organic part of the structure ; but its
characteristic quality is subordinated to its function as an
invisible support for drapery. Naturalism, as a deliberate aim in
art, became a disease in the South, when, like decadent Greece,
it began to think in terms of science.
in. Mazdean Ideas in Christian Representation
Up to this point my treatment of representational art has
been based upon known facts ; only my interpretation of the
links in its development rests upon impressions formed by the
experience of many years. But there is one province, the most
important in the whole range of church-art, which I have so far
left untouched. Its peculiar and dominating feature has never
been adequately emphasized or defined. I refer to the mosaic-
decoration of the apse. In this vaulted ending of the church
towards the East, invariably present in the wooden-roofed basilica,
a form of decoration was developed which, if I am not mistaken,
has had an extraordinarily widespread influence on art. Rome
and Ravenna are the only places in which examples of this style
have survived ; we shall therefore begin by considering the well-
known monuments on the Tiber and the Adriatic.
The mosaics of Rome and Ravenna show clear traces of
a non-representational substratum, which was gradually super-
seded by representation. The apses of S. Maria Maggiore
and S. Giovanni in Laterano, the last great survivors of their
kind, sufl'ered much damage at the hands of Torriti about
A. D. 1300, but not so much as to preclude us from reconstructing
their original features. There is no need to discuss this subject
again. But we must give our attention to the gradual rise of
representation, which can best be observed in Rome and Ravenna,
and in apses extant elsewhere. The oldest of these mosaics, that
of S. Pudenziana, built under Innocent I (a. d. 402-17), is an
excellent example of the ecclesiastical style ; it is the most
important outcome of the representational stream which flowed
from Nisibis-Edessa-Antioch for the difl^usion of Christian
172 TRIUMPH OF REPRESENTATIONAL ART
doctrine. In Rome, as in Ravenna, it was probably not before
the fifth century that the full force of this Eastern influence made
itself felt. I question the recent theory to the effect that S. Puden-
ziana shows indisputable influence from Jerusalem. The cross
in the apse conforms to the general Eastern usage. Here, as in
the mosaic of S. Giovanni in Laterano, it stands upon a hill.
The scene depicted below, including the figures of Christ and the
Apostles, is merely an elaboration of the Heavenly City, which,
in the Lateran mosaic, is seen on a miniature scale in the hill
under the cross. The resemblance in the grouping of the figures
of Christ and the Apostles in S. Pudenziana and the Antiochene
Pyxis in Berlin can hardly be accidental. The architectural
background is modified to suit the requirements of a monumental
art, and opens out in semicircular form, as opposed to the type
seen in S. Prassede, where the Heavenly City is enclosed by
a wall.
Hitherto we have divided the representational art of the
Christian Church into two distinct provinces : one, the purely
Hellenistic province of Asia Minor, the symbolism of which,
in Alexandria, admits a partly Jewish strain ; the other, the
province of Antioch, the emporium of the Semitic centre of
Edessa-Nisibis, where symbolism is replaced by didactic purpose.
The latter had an ever-increasing range of influence in the West,
Jerusalem and Constantinople acting as agents of diffusion. It
still remains to call attention, in the representational art of Italian
mosaics, to an older influence which, in my opinion, must be
connected with Iran, and more particularly with Mazdean ideas.
It is a striking fact that mosaic, the decorative covering of vaulted
surfaces, should be the chosen medium for the transmission of
this influence ; its use may perhaps be regarded as a clue
pointing to Northern Iran, just as ivory, in another field, gives
a clue with regard to India. Before entering upon a discussion
of these mosaics, it may be well to repeat certain pertinent
observations which I have already made in my work on Armenia.
By way of a general preface we should note that Zoroas-
trianism adopted an attitude antagonistic to the ancient classical
world, developing along thoroughly East-Aryan lines ; later, at
the time of the Reformation, it strove for recognition also in the
North within the pale of the Southern religion. Mazdaism
preaches in its purity a moral conception of the world as divided
between Good and Evil. The protagonists of Classical Christian
HELLENISM, SEMITISM, MAZDAISM 173
Art have paid too little attention to this fundamental characteristic.
It was left to the Sassanian dynasty to make the almost Christian
teaching of Mazdaism subserve the interests of property and power
in the spirit of the older Semitic and Mediterranean cultures.
Before Sassanian times Mazdaism had no need of ' representation';
this fact in itself, quite apart from the perishable nature of
unbumed bricks and stucco, is sufficient to account for the
absence of permanent monuments, and the unhampered con-
tinuance of the decorative art characteristic of Northern and
nomadic peoples. Two other Aryan religions, the Buddhist and
the Greek, in like manner created forms of art untrammelled
by ideas of power and property, though in their case the human
figure was included ; only after the death of Asoka and Alexander
respectively did they lose their essentially Aryan character
through their adoption by imperialist religions. Ancient Iranian
art, in this perhaps following in the footsteps of old Aryan
religious ideas, made no representation of God ; even in later
times it only did so rarely, and then, probably, under foreign
inspiration. Attempts to represent God in Achaemenian times
may be fairly ignored, since they had no influence upon the later
popular art, with which we are here concerned. How it was
that Zoroastrianism failed to obtain any influence over the
Achaemenian court is a question which we are not yet in a position
to answer. The motive of the mounted God will be examined
presently. The hypothetical figures of God and Zoroaster on
one of the reliefs of Tak-i Bostan belong to Sassanian Court art,
and are not here in point. There can be no question of a popular
religious representation. It is the art of a royal court that has
need of visible figures, in order to make manifest a power con-
ferred by the grace of God. First Assyrian-Babylonian, subse-
quently Hellenistic art supplied Persia with figures for that
purpose. Indian and East Asiatic influences only intruded into
Early Christian art in so far as they accompanied raw materials
such as ivory and silk.
A. The Mounted Saint. The contest between Good and
Evil is a subject familiar to students of Christian art in the East
through representations of the Mounted Saint. The two pro-
blems which it is our present purpose to investigate are, first,
the date and place at which the God was originally depicted as
an equestrian figure, and, secondly, the date at which this figure
developed into a champion of Good against Evil.
174 TRIUMPH OF REPRESENTATIONAL ART
In the earliest times the Aryans were not horsemen, but
drivers of carts. But we already find them mounted in the
Rigveda ; and the game of chess has preserved ocular proof of
this up to the present day in the figure of the ' knight ', just as
the ' castle ' recalls the elephant and the ' bishop ' the wheeled
vehicle. In the religious ideas of Mazdaism the six great good
spirits, the Amesha Spenta, are represented as mounted figures.
This conception runs through the whole Avesta ; I will give
only a single instance. In the Zarathusht-Nameh (a work
written for Parsis in New Persian, and completed in a. D. 1278)
we read how two of these Amesha Spenta (good spirits or perhaps
angels) appeared, with two holy fires, before King Gushtasp
in the form of mounted figures (Suvaran), carrying weapons of
war. Each was like a walking mountain, clad in warlike dress
and coat of mail. . . . All of them wore the sacred colour of green,
the colour of angels, and were equipped with weapons : ' they
' brandished their lances before the King . . . each sitting thus
* upon his horse.'
Concrete evidence for the mounted figure of God in art
may be seen in the Sassanian rock-cut reliefs which depict the
investiture of the ruler by means of two confronted figures on
horseback, representing Ahuramazda and the prince. The horse
was so integral a feature of Aryan life that it is inseparable
from the Aryan idea of God.
But there is a much more convincing piece of evidence
which I have had in mind for years. In Egypt in 1901 I first
discovered that not only S. George, but every Saint, even Christ
Himself, was represented as a mounted figure (Fig. 6o,facing p. 166).
Even at that time I assumed that for the origin of the type ' all the
' signs pointed in one direction, towards Iran, whence the horse
' itself was introduced into the West '. The equestrian relief from
Suweida in Syria gave me the opportunity of expanding this
hypothesis, for which so much evidence has since accumulated
that it has gradually developed into a certainty. First of all
a whole series of such figures of mounted saints was discovered
in Christian Mesopotamia ; later I found them widely distributed
in Armenia, one such figure representing the Founder, and others
as at Achthamar a large variety of saints. The fundamental idea
is clearly that of the saint as horseman, an idea demonstrably
known to Mazdaism. Whether the popular religion of Iran
actually translated the idea into representational form, or whether
HELLENISM, SEMITISM, MAZDAISM 175
this was first done through the medium of Hellenism on Persian
soil, as in the case of Court art, is another question.
So far we have only considered the horseman and his
equipment. But I am inclined also to attribute to Mazdaism the
conception of this figure as a champion against evil. Formative
art carries us farther than legend in this matter. It is significant
that the Alexander of the Pompeii mosaics, the Heracles Maxi-
mianus of Suweida, and the Horus of the Louvre are all depicted
as armed and carrying the lance in the manner of the Amesha
Spenta. Like these they should properly be shown striking down
an enemy. There seems to be evidence to prove that this was
a favourite subject in the symbolic decoration of church gates
in post-Constantinian times. In my published work I have often
had occasion to discuss the mosque gate of Dashlut (Fig. 54,
facing p. 152), where the tympanum has apparently been trans-
ferred from a Christian church. Actual church gates of this type,
with representations of God as a mounted warrior, may not exist,
but they are attested to by Syrian literature. An Arab tradition
ascribed by older commentators on the Koran to Mohammed
himself relates the fact in the case of Lydda. Even Christ is
there said to have been depicted mounted on a mare and slaying
Anti-Christ. The same commentators render it probable that
the slaying of a wild boar adorned in like manner a gateway
in Jerusalem. The practice is also corroborated by the narrative
of Eusebius, who relates that Constantine had his own figure
painted over the entrance to the imperial palace, with the Cross
over his head and a dragon under his feet. Eusebius explains
this as another method of suggesting the victory of the Faith —
a favourite subject in the Egypt of the fourth century,' In this
particular case, where the victory is symbolized by a mounted
figure slaying a beast, the type may well go back to the Mazdean
conception of the triumph of Good over Evil.
The same subject reappears constantly in Armenia. An
exhaustive search commencing in that region would lead us straight
back to Iran. In the representation of the mounted saint, Egypt
has as usual retained the original Irano-Armenian character in
its purest form, as is the case with the mounted figure of Christ,
which I brought to the Berlin Museum (Fig, 60, facmg p. 166). In
the West the same idea survives in the form of S. George, and
attains a lofty artistic expression in Diirer's * Knight, Death, and
^ Cf. the wood-carving in the Kaiser Friedrich Museum at Berlin.
176 TRIUMPH OF REPRESENTATIONAL ART
the Devil '. In this arresting picture the German master almost
seems to have had intuitive knowledge of the ancient Aryan
conception, and to have applied it at the time of the Reformation
to the idea of the Christian knight.
B. The Last Things. In treating of the miniatures of the
Serbian Psalter in the National Library at Munich, I had occasion
to discuss three unusual subjects, * The Cup of Death ', ' The
Tree of Life ', and * The Uncovering of the Bones '. The second
of these subjects derives from the legend of Barlaam and Joasaph,
and ultimately from Indian sources ; a similar pedigree recurs
in the ' Physiologus '. The struggle for the soul is a thoroughly
Mazdean idea. The best incentive to a closer study of this
subject is afforded by the motive which our familiarity with
the Campo Santo fresco at Pisa has led us to describe as the
Triumph of Death. Other students must have been impressed
by the Oriental derivation of certain features in these pictures.
The apocalyptic cycle likewise suggests a Mazdean background.
C. The Last Judgement. The setting of the Last Judgement
scene at Pisa closely approximates to Mazdean Christian ideas.
It is the consummation of those Last Things which are among
the oldest traditions of the Avesta. It accords so perfectly in
every part with Iranian and Jewish conceptions, that by turning
to the work of Ephraim the Syrian {d. a. d. 373), a pupil of Jacob
of Nisibis, we can find a summary of all these subjects, and
follow their development in an unbroken sequence down to
the time of Dante. These relations must be realized, if we would
understand how such a profusion of detail came to exist as early
as the fourth century, and how both the converted Bulgarians
and Mohammed were able to make such full and efficient use
of them. All this mass of details was already present in
Mazdean conceptions before the appearance of Christian art,
which merely added the representational element. To illustrate
the importance of Mazdean influence in the apse mosaics of
Rome and Ravenna, I shall examine one example m each of these
places. For the study of the subject of the Last Judgement in
general the reader is referred to the large representations on the
west walls of church interiors, such as the mosaic at Torcello
and countless other examples in orthodox churches. I would
suggest a comparison of these with the text of Ephraim and with
the ideas of Iranian and Jewish eschatology. My present
purpose is to trace the pure symbolism of fourth-century art to
I
HELLENISM, SEMITISM, MAZDAISM 177
its source, and to show that the idea of the Last Judgement
already dominated the art of mosaic, in a form not yet recognized,
before its transmission from East to West. The artistic influence
of Antioch is relatively of minor importance.
In the preceding chapter I showed that Early Christian art
made use of a kind of landscape with fishing and hunting scenes,
scrolls of vine and pomegranate, and a rich symbolism based on
animal and plant life ; and that certain remarkable features of
this art point to Iran as the source from which its underlying
ideas were derived. I shall now attempt to supply the key to
its interpretation, in so far as what has been said about Hvarenah
has not already done so.
We may begin by examining one of the smaller and more
unassuming mosaics in S. Apollmare Nuovo, along the top of
the wall above the windows. The reader is probably aware that
their subjects are arranged with figures of the prophets as we find
them in the Syrian Bible from the Mesopotamian monastery of
Zagba, and that the choice and sequence of the biblical scenes
agree in general with Syrian liturgy. An Easter lection from the
Gospels akin to that used by the Jacobites has here been pictorially
translated. Let us take the mosaic depicting the judgement of
the sheep and goats. Christ is there shown sitting amidst a rocky
landscape accompanied by a red and a blue angel and stretching
forth His right hand towards three white sheep, opposite which
are three black-spotted goats. Paulinus of Nola, who also refers
to the Hvarenah landscape, describes the subject of the goats
and sheep in an apse mosaic of the Basilica at Fundi. It depicted
a broad landscape, in the centre of which stood the solitary
figure of Christ on the rock ; it thus differed somewhat from the
well-known mosaic over the entrance to the Mausoleum of
Galla Placidia, which belongs to the same class. The Shepherd
was shown caressing four sheep at his right hand, and motioning
four goats away at his left. Above the figure of Christ were the
symbols of the Trinity.
The allegory of the sheep is also used by Ephraim the
Syrian in his Last Judgement, and appears, contemporaneously
perhaps, on the lid of a Roman sarcophagus in reference to
S. Matthew xxv, 32. The idea in itself was very probably
Mazdean in origin, and is described in the Avesta. At the
resurrection of the body, we read, every one will see his good and
his evil deeds. ' Then shall be held the assembly of Catva9tran
245 » N
178 TRIUMPH OF REPRESENTATIONAL ART
* and the Evil shall be separated from the Good as the black sheep
' from the white.' I have already pointed out that the figure of
Christ as the Good Shepherd is a Mazdean conception (Yima) ;
the mosaic in the Mausoleum depicts Him in a landscape of the
Hvarenah type, and the sheep in Early Christian mosaics indicate
a similar origin,
D. The mosaic of SS. Cosmos and Damian (Fig. 63, facing
p. 180). I shall now consider in some detail the mosaic of
SS. Cosmas and Damian, artistically the finest of the Roman
mosaics; it holds a unique position, outside the ordinary lines of
development on Roman soil, and similar to that occupied in the
architectural field by the * Minerva Medica '. The mosaic depicts
red clouds sinking towards a band of water behind the land in the
foreground, and filling the whole central portion of the semicircle.
Palms on either side, a phoenix in one on the left, carry the eye
upwards to other clouds floating at the summit of the apse. This
' landscape ' on a dark blue ground reveals the same composite
symbolism of clouds, water and earth, phoenix and palm, which
I have already explained as typically Mazdean. In the midst of
this setting appear Christian figures ; in the centre the Saviour,
with large gesture, standing high up on the clouds ; on either
side smaller figures of the saints led in by Peter and Paul ;
finally the ' Founders ' ; the whole somewhat theatrical com-
position showing a calculated effort to impress the spectator.
When I consult published works on Roman mosaics with
a view to discovering what is meant by this creation of Pope
Felix IV (526-30), I mil to find any appreciation of the landscape
as such, but, at most, only a passing reference. Wilpert is of the
opinion that the ' Jordan ', flowing behind the figures, is rather
meaningless in this position and might equally well have been
omitted. He frankly describes the mosaic as the Abode of the
Blessed or the Heavenly Paradise, and considers that in the
figure of Christ the artist intended to emphasize the divine
nature. That is why the artist has depicted Him standing upon
clouds as though He had just appeared, and the light which
radiates from His form reddens the clouds like the rising sun.
The apostles, he goes on to say, are intended to draw attention
to His presence ; the martyrs are in the act of turning towards
Him. The blue ground heightens the mystic character of the
whole scene, while the gold of the lower zone with the sheep
accords better with the more joyous character of Christian
HELLENISM, SEMITISM, MAZDAISM 179
symbolism. That is all that the Roman specialist is able to say
about this work of art. Other authorities have succeeded in
advancing a stage farther. Zimmermann is right in laying stress
on its essentially representational character, and on its solemnity,
which inspires in the beholder an involuntary feeling of awe and
silent reverence. Let us go more closely into the matter.
This mosaic stands outside the series of fourth-century
mosaics, whether decorative or depicting only landscape ; nor
can it be included among those of the fifth century, which
represent Christ as the Teacher among the apostles. Yet Wilpert
considers that we should be mistaken in regarding it as a wholly
independent work of art and the first successful attempt at apse-
decoration. There is evidence, he says, that earlier models were
available ; this is shown by the account of the apse-mosaic of
S. Lorenzo in Lucina, which was at least a century older and
showed a strong resemblance to it. Wherein lay this resem-
blance ? In the landscape ? That cannot be ; for it depicted
wreaths and foliations, as in S. Maria Maggiore, and single
figures. Are the figures of Christ, with Peter and Paul, the
protomartyrs and the founders, really an original part of the
composition ? And are the human figures really of primary
importance from the artistic point of view ? Clearly, the great
feature in SS. Cosmas and Damian is the boldly conceived
landscape consisting of earth, water, and clouds arranged in
a series of distinct masses. The scroll designs of S. Lorenzo in
Lucina have not the remotest connexion with this form of land-
scape ; they are more nearly related to the mosaics of S. Maria
Maggiore and the Lateran. But it is precisely the landscape
which distinguishes the mosaic. I am not concerned to trace
the figures of ' founders ' to the Church of S. Peter and the time
of Constantine, or to show that the figure of the summoning
Christ is derived from some pre-Constantinian Roman church
art, imitated in the Catacomb of Domitilla. My object is to
grasp the full significance of the pregnant form of art employed
by Pope Felix, and to probe it to its source.
We may learn from inscriptions on Armenian churches that
they were sometimes founded for an intercessional purpose ; this
is corroborated by inscriptions on tombstones. One of these
runs as follows : ' Mayest Thou (Christ), when thou appearest in
majesty, intercede for me, an unworthy servant of Christ.* The
inscriptions of SS. Cosmas and Damian run similarly : Aula Di
K2
i8o TRIUMPH OF REPRESENTATIONAL ART
Claris radiat speciosa metalUs \ in qua plus fidei lux pretiosa micat \
martyribus medicis populo spes certa salutis \ venit et ex sacro crevit
honore locus._ This is followed by the dedicatory inscription :
optuUt hoc Dno Felix antistite dignum \ munus ut aetheria vivat in
arce poli. (* Felix offered unto the Lord a gift worthy of a Pope,
in order that he might dwell in the Heavenly citadel '.)
The event depicted in this mosaic is the great Epiphany of
Christ, of which Ephraim the Syrian in his description of the Last
Judgement says : * Behold, the day of the Lord breaks suddenly
* upon Creation and the righteous draw near to him with burning
* lamps (represented in this particular case by the crowns of
' the Martyrs). But I am in utter darkness and have no oil in
* my lamp, wherewith to go and meet the bridegroom when
' he comes.' This is the true explanation of the mosaic. The
conception is reminiscent of the Avesta and that belief in im-
mortality and eternal judgement which was a chief dogma of
Mazdaism from the earliest times. The twenty-second Yasht
describes the fate of souls after death ; it also contains a passage
which supplies the key to the landscape of SS. Cosmas and
Damian in words that almost literally describe it : * At the
* breaking of the fourth day, and the rising of the rosy flames
* of dawn, when the gates of Heaven are opened ' the soul goes to
judgement. And at the Last Judgement, according to Yasht 19,
the Redeemer of the World is supposed to rise from the water
Kansu, * from far away out of the East, the original source and
* dwelling place of the light. It is his task to accomplish the
' renewal of the world. He maketh the living immortal, and the
* dead he awakeneth out of their sleep. He putteth an end to
' age, death, and decay. To the godly man he giveth everlasting
' life, everlasting happiness and the fulfilment of all his desires.'
This is the Iranian conception, to which the mosaic of Pope
Felix gives visible expression ; the treatment of the figures is
its oruy Graeco- Semitic element. The introduction of the
figures of the * founders ' is alone sufficient to indicate an eastern
origin. But the landscape, which would be more effective without
the human figures, is clearly derived from Iran. The * Jordan '
is not a negligible element ; it is precisely the water, the ocean
of the world, which, in conjunction with earth and clouds,
forms the decisive factor. The religious landscape transcends in
importance the human figures with their calculated dramatic
effect, and gives to the picture that air of simple and sublime
00
N
00
a,
a,
u
lU
a.
C3
to
o
C
Q
•a
c
C4
O
U
(/!
iT
S
o
OS
64 Kioto, Zenrinji temple ; figured silk ; after the /Co^Aia. See p. 181.
HELLENISM, SEMITISM, MAZDAISM i8i
repose, which characterizes not Greek art alone, but also, in its
distinctive fashion, the art of Eastern Iran.
It was my intention in this volume to deal exclusively with
Christian art ; but I should like to make an exception at this
point in order to point out how great and alluring is the goal that
awaits those who apply the comparative method to art. Phenomena
very similar to those which we have noticed in Roman mosaics
occur in a Buddhist silk picture of the Zenrinji temple at Kioto in
Japan (fig. 64, opposite). In this picture the figure of Buddha,
the religious founder, represents the dawn, and is accompanied,
in place of Peter and Paul, by his Bodhisattvas Kwannon and
Seishi, beneath whom are smaller figures of the four Deva kings
and two Founders. The essential point is the use of landscape
to introduce cosmic ideas. Buddhism, in its passage from India
to China through the Mazdean culture area, seems to have had
its ideas modified much in the same way as Christian art. For
if I am not mistaken, the picture in question is a reproduction
by a Japanese in the thirteenth century of an original work by
the great T'ang master artist Wu-Tao-tse who lived in the
eighth. Curiously enough, the Zenrinji picture has been eulogized
n practically the same words as the mosaic in SS. Cosmas and
Damian, though less with reference to its subject-matter than to
ts artistic worth. I quote the words of Otto Fischer, which run
as follows : ' The Amida of the Zenrinji picture, rising like the
moon from behind the hills, is one of the most wonderful con-
ceptions in the whole range of religious art. The theme is treated
in a spirit infinitely serene like that of autumn, and permeated
with such an atmosphere of celestial peace, that the sight of it is
one of the most impressive of all experiences. Yet here again
this unique effect is the result of an abstract composition and
partition of space, which depend upon conditions admitting of
calculation, symmetry and gold contours, pyramidal construction,
the balance of curves, and the skilful variation of scale ; such were
the means which the painter employed. They give the picture,
notwithstanding its inspiration, a sense of destiny detached from
the emotions of the human kind. Even the wave-like outlines
of the hills fringed with trees do not resemble those of experience
or sight, but are disposed as decoration according to the laws of
ascending and descending curves, so as to form an effective setting
for the symmetrically inclined figures of the attendants, and the
supreme repose of the facing Buddha.' If we apply this subtle
i82 TRIUMPH OF REPRESENTATIONAL ART
analysis to the Christian mosaic at Rome, bearing in mind my
repeated opposition throughout this work of Iranian form to
Hellenistic- Semitic objectivity, we shall perhaps get some idea of
the purifying influence of Mazdaism upon art. The truth is indeed
as little recognized in Japan as in Europe. The picture at Kioto
is generally supposed to be connected with a vision of the priest
Eshin Sozu.^
With our horizon enlarged by comparison, we are in a better
position to appreciate the narrowness of the classicists, whose
attitude is well illustrated by such a statement as the following :
' If there still exists anything in Christian art for which we cannot
* point to a pagan, that is a classical analogy, we should remember
* that this may at any moment be supplied by a new discovery or
* excavation.' We know that the Greek spirit, in its earliest period,
was far too robust to think of explaining this world by reference
to a future state. The change of view began with Plato, and was
probably not wholly uninfluenced by the East, where an Aryan
religion centring round this dream had arisen two centuries
before. In the origin of Christian art, Mazdean influence upon
Graeco-Roman ideas was not merely indirect ; unless I am
mistaken, Zoroastrianism, which at the birth of Christianity was
still at the height of its power, exercised a direct and not incon-
siderable influence on Western thought. One-sided insistence on
the essentially classical character of Early Christian art is bound to
provoke counter-movements with such watchwords as ' Christian
Semitism ' and * Christian Mazdaism '. Classical Christian art
dies out in the course of the centuries ; Christian Semitism
remains triumphant in the field of representation, and Christian
Mazdaism in that of architecture. With these West- Aryan
classicalism intermingled in Byzantium, and later also in the
West, but only at sporadic moments of revival.
IV. Influence oj the Court
The object of Edessa-Nisibis was didactic. Byzantium, on
the other hand, made church and religion subservient to its im-
perialist policy and, like the Sassanians, transfused all things with
^ The subject here represented is of Amida, but that up to the present no
Yamagoshi-Amida. H. Schmidt kindly evidence is forthcoming for the earlier
informs me that he is inclined to connect treatment of Yamagoshi-Amida or the
the type with the Taema-Mendara of the influence which it had on later art.
eighth century, and with Eshin's studies
HELLENISM, SEMITISM, MAZDAISM 183
the spirit of the court. Tangible proof of this is afforded by
the manner of depicting the Virgin and her Hfe. The most
characteristic examples preserved are the apse-mosaic of Parenzo
and the Adoration by two processions of Saints (male and female)
in S. ApoUinare Nuovo at Ravenna, in both of which examples
the royal treatment of the Virgin is derived from the church of
the Nativity in Bethlehem. In both cases she appears as an
empress enthroned, and separated from the Saints and founders
by attendant angels. This spirit and the source whence it sprang
are clearly shown in the dogmatic mosaics on the triumphal
arch of S. Maria Maggiore, and still more in the choir of S. Vitale,
where, appropriately enough, a representation of the court with
all its pageantry is bodily introduced. This court style sub-
sequently gives way in some degree to the purely ecclesiastical
treatment of Edessa-Nisibis, although a special change takes
place in apse-decoration by the substitution of the Virgin for the
Cross.
In conclusion I must clearly repeat that I have ignored the
view of Christianity as a faith which created an art out of its own
resources, except in so far as an art relying on intellectual content
was encouraged by the theologians, or, to come down to matters
of detail, church architecture was made to conform more closely
to the new liturgy by the introduction of the apse and the con-
nexion with it of a longitudinal axis, and other changes of this
kind. But, to be candid, we are only at the threshold of serious
research. I therefore feel it necessary at the outset to dissociate
myself from any standpoint restricted by a confessional point of
view or by a narrow horizon such as that which ascribes all
initiative to Rome. There are some who can see nothing but the
Church and ' representation ', and imagine both essential to
religious experience in the field of art. Yet it is important, even
in considering the Early Christian art of the large Hellenistic towns,
and among them Rome and Byzantium, to have a clear idea how
the Church began, and when representation first arose. It must
be definitely realized that neither Christianity nor representation
originally had the slightest connexion with Rome or its church
architecture, pregnant though that was with possibilities of develop-
ment ; we should further remember that the Hellenistic timber-
roofed basilica to which Rome, once the home of a most brilliant
vaulted architecture, clung with a strange persistence, began in
course of time to yield to the pressure of Eastern communities,
i84 TRIUMPH OF REPRESENTATIONAL ART
while the vault, gradually advancing westwards, first profoundly
modified the long church, and ultimately, after a thousand years,
sealed its triumph with the dome of S. Peter's, Meanwhile both
court and church vigorously collaborated in the destruction of the
peculiarly Northern style known as Gothic.
V. The West
Representation established itself in Western church art
mainly in the regions dominated by the South ; the retreat of
representation is m itself the proof of a strong indigenous feeling
for art. The Goths and Lombards had no representational art
at the time of their invasion of Italy. The crude figures which
they subsequently produced bear in iconography the sign-
manual of Byzantium, but in their form are Northern ; they also
show an East- Aryan strain, particularly in their use of symbols,
which at times are quite reminiscent of Hvarenah decoration.
This is perhaps even more marked in the Balkan peninsula, in con-
temporary Greece, and the old Croatian empire, where the dislike
of representation manifests itself either in the modification of
classical forms or in a faithful adherence to Byzantine, that is,
Aramaean types ; these, however, retain the treatment of figure-
surface with triple-grooved bands after the manner favoured in
Armenia and in Ireland.
The contact of the Prankish empire with the East is clearly
revealed in illuminations of manuscripts, but Charlemagne em-
Eloyed the services of the Greeks and Syrians for other purposes
esides that of editing the gospels. Architecture reveals, both in
structure and decoration, connexion both with contemporary
Eastern forms and with the old Hellenistic basilica. The occa-
sional use of classical models and a certain hesitancy between
the representational and non-representational styles is observable
throughout the whole range of art, including illumination.
Decorative and representational art go hand in hand, and it is
rare to find even a tentative efl^ort at something truly indigenous.
Both subjects and figures were borrowed from the most diverse
sources ; it is in the treatment of form that the resistance of the
North to the victorious church comes out most strongly. As
in Armenia, there is a tendency more or less marked to modify the
lines of the newly adopted subjects ; at times they receive the sug-
gestion of unrestrained movement and become true vehicles of
HELLENISM, SEMITISM, MAZDAISM 185
expression. The same tendency appears in the best period of
Romanesque ; we may compare the sculpture of Moissac with
that of Armenia.
This formal tendency reached the highest level of accomplish-
ment as soon as the North began to manipulate the forms evolved
independently in the South-West. A logical development ensued,
which for vitality of form can only be compared with the achieve-
ments of Greece and of Armenia. The result was a lofty expres-
sionist art, unspoiled by scientific naturalism and in harmony
with the spirit informing the structure as an organic whole. It is
thus entitled to be called great art in the best sense of the term.
Herein the Northern (Gothic) cathedrals attained a height hardly
surpassed by Donatello, in whom, indeed, we mark an attempt to
study classical sculpture and natural forms together. This was
an unfortunate attempt repeated by many painters, by Raphael
not least ; to it we owe in no small degree the subsequent con-
fusion between art and science which has since but too often
characterized the work of masters with individuality and of the
unoriginal herd. Only the greatest were able to avoid the pitfalls
of naturalism, intent on its near objective ; only they could make
representation express their own souls and those of their peoples
in a manner so exalted that even in the strait limits of the pictorial
they remained great artists. Leonardo, Michelangelo, Giorgione,
Diirer, and Rembrandt can never be surpassed in this kind.
The North preserved its own genius in representational art,
as the East-Aryans of Asia theirs in landscape and in the expres-
sion which they achieved by abandoning imitation of Nature.
In the end the South too was swept away in its turn by this
Northern torrent. The spiritual quality which Diirer infused
into a hackneyed subject in his picture of ' All Saints ', with its
rosy dawn and the clouds sinking low over the seascape, was also
profoundly felt by Goethe. When he says that the sun setting
over the sea will always be the most sublime of symbols, we are
reminded of the Hvarenah landscape in Mazdean art. The
classic becomes a romantic when he adds : ' Even in its setting it
abides the same.'
The origin of Christian Church art is much more definitely
interlinked with geographical, racial, and national characters than
people have hitherto been led to suppose ; they have under-
estimated the importance of the list of peoples mentioned in the
Acts of the Apostles as present in Jerusalem at the first Pentecost.
i86 TRIUMPH OF REPRESENTATIONAL ART
Spiritual movements are superficial in comparison with the fixity
of geographical factors ; they may modify this or that, embrace
foreign influences, and even end by losing their identity ; but
they can never produce new or decisive values. Christianity and
its art had no true roots of its own ; wherever it went it stayed
itself upon the local and national foundations which it found in
existence during the first four centuries of our era. By following
the list in Acts ii. O-n we can observe how the Jewish nation
spun the threads of its distribution not, at first, in the direction
of Asia Minor or Rome, but towards the lands of the Parthians,
Medes, Elamites, and Mesopotamians, the very peoples, in fact,
who created vaulted buildings and gave their walls that decorative
lining, the object of which was not to represent but to adorn.
The Jews, who had no representational art, delivered infant
Christianity to Mazdaism at a time when the Persian people had
won control of their art and the Sassanian dynasty had not yet per-
verted it into a state religion. The mosaics of Rome and Ravenna
afford the clearest proof in a style of decorative Hning specially
designed for the vault. In historical interest the churches of
Armenia are almost surpassed by this species of art, which was
particularly applied to the decoration of the apse in the timber-
roofed basilica. The purely ornamental and landscape forms of
decoration introduced from Iran still formed an attractive sub-
stratum in the art of Rome and Ravenna at a time when Semitic
figure-subjects began to predominate. The mosaics of SS. Cosmas
and Damian and of S. Apollinare in Classe are admittedly ana-
chronisms in the sixth century. At that time Semitism and Hellen-
ism had already contracted the alliance which remained their
fixed policy, an alliance to which Iran, for all the efi"orts of the
iconoclasts, offered an ever- weakening resistance.
In considering all this interaction between countries, races,
peoples, and spiritual forces, we cannot leave India outside our
range of vision. In the province of formative art we find links
which show Gandhara in its treatment of Buddha and his story
so nearly related to Asia Minor and the home of Semitic Chris-
tianity that the discovery of similar affinities in ivory carvings no
longer excites any surprise. This connexion explains the transi-
tion from Indian symbolism to the historical style of Gandhara.
The connexion of Christian art with India was especially strong
in Egypt, where it is perceptible even at the present day.
It will therefore be necessary to broaden the old lines of
HELLENISM, SEMITISM, MAZDAISM 187
research : the new school can no longer confine its view of
Christianity to Greece and Rome, but will extend its horizon
so as to include Hellenistic Asia, the Semitic countries and Iran.
As a general observation it is true to. say that the proper creative
period of Christian art terminates with the end of the fourth
century ; the letter of Nilus clearly indicates the change from
the unsophisticated style of Eastern communities to one according
with the strict demands of the church in the Mediterranean area.
It was only in the north of Europe that fresh creative energy
survived, to produce, in the twelfth century, that flower of
Christian art which we call Gothic. This art developed in the
north of France out of a lost wooden architecture, just as
Armenian construction was derived from vanished Iranian build-
ings of unburned brick, though in either case a considerable
previous development in stone or in rubble building must be
presumed. It will require generations of research to guide these
streams from their first sources, which I have been able to indicate,
into their broad and true channels, and to follow up all those
divergences and creative innovations which lend such richness to
the total picture. But the mere ascertainment of such historical
facts is not our ultimate object ; these facts, critically examined
and sifted, must themselves serve, after their kinds, as a basis for
comparative research. It is only at this point that the task of the
specialist can be taken up, and a systematic history of develop-
ment begun.
In the present volume I have repeated the attempt, so
frequently made in the last decades, to fill up one of those gaps
which have hitherto made it impossible to carry out any work on
Early Christian art or on the emergence of the Middle Ages from
the evolutionary point of view. First Byzantium and the great
Hellenistic cities, then Asia Minor, Armenia, and the group
formed by Edessa, Amida, and Nisibis in Northern Mesopotamia,
and now finally Iran, — all in turn claimed our whole attention.
But do they really complete the circle ? Are no gaps still left to
fill ? On the borders of this enlarged horizon lie India and Eastern
Asia. The country beyond the Oxus, between Altai and Iran, was
as it were the clearing-house whence routes led northwards as
far as Scandinavia, and southwards to the great emporia of Syria.
The Indian origin of the fish-symbol has formed a subject of dis-
pute ; Buddhist temples in Turkestan have been claimed as inter-
mediaries by those in favour of the theory, and rejected by its
i88 TRIUMPH OF REPRESENTATIONAL ART
opponents. But such trifling details lead nowhere. Mazdaism
and Buddhism already existed in a high state of development
when Christianity forced its way into their territory. This is a
fact which can no longer be ignored in the reckoning with Hellen-
ism as now represented by classical philologists and archaeologists.
But what do we know about the remains bevond the Oxus r Or
indeed about those of China ? Strange surprises await us in these
regions, more especially in connexion with the Italian Quattro-
cento. This investigation into the origin of Christian Church art
(from which sepulchral art has been purposely omitted) has led
me, in the preceding historical section, to what has been called
a complete transvaluation of all existing values. The reason is
that I have not, like a philologist, been confined to a single language,
or pursued research in the interest of a single and sharply defined
province. It has been my advantage to employ a tongue universally
understood, the lingua franca of formative art ; I have thus been
able to roam freely throughout the territories of Hither Asia and
so to comprehend the whole. In the old days special weight was
given to Latin sources, more recently to the Greek and the
Syrian as well. To-day the student can bridge the gap left by
Pehlevi,and profit by the documents of Chinese Turkestan which
are now being deciphered. But the most important advance of all
came from the discovery that Armenia and the country of the
Arsacid dynasty which gave it Christianity belong to the North,
^/\^ ^ and may be held to have effected a penetration of Southern
VHither Asia and its culture almost without a parallel.
The Arabs would never have raised Islam, nor the Jews
Christianity, to those spiritual heights which they attained at
any rate in the realm of art, had not the Iranians, living under
similar economic conditions, supplied both movements with the
vigorous style created by their national Mazdaism. The problem
of the origin of Christian art cannot be solved by assuming
spontaneous evolution from within, and accepting in a general way
the significance of Hegel's ' content '. Rightly to appreciate the
development of art in Early Christian times, the student must
embrace in a comprehensive view peoples and races, the material
used and the purpose conceived ; he must know how clearly to
distinguish the essential values of formative art. I shall therefore
round off this essay in development by a formal survey of the
results yielded by this wide outlook in the purely artistic field.
VIII
Systematic Investigation of Essential
Character and Application of
the Comparative Method
IN the preceding chapters we frequently had occasion to allude
to a certain method in the study of the history of art, a method
which seeks not merely to describe and sift monuments and
sources, but also to ascertain and interpret the course of their
development. This method of study, which I may describe as
research into essential character, is fundamental to our subject.
While historical investigation precedes it and helps to prepare its
material, its own scope is limited to the work of art itself. It thus
forms a necessary preliminary to a third kind of research, which
deals with the history of development in art, and is closely
associated with the study of various other aspects of life, including
not merely music and literature, but such matters as power and
property, law and custom, economy and technology. These are
the aspects and activities of life with which formative art comes
into contact ; and scientific workers in these fields would readily
open up relations with the art student were they not precluded
from fruitful co-operation by the narrow horizon of contemporary
oflicials and professors of art, who are too comfortably involved
in their own small European field to be capable of devoting serious
attention to things over its frontier. Some one place there should be
where the whole range of problems concerning formative art might
be conscientiously kept under review ; but so long as there are no
institutes of research served by trained experts with adequate
remuneration, the gap must be filled by the institutes of art at our
universities, attracting audiences from all the various ' faculties '
and striving to inspire in them an interest that shall outlast student
days.
What I call the investigation of ' essential character ' builds
on a groundwork of historical fact and has no connexion with
iQo INVESTIGATION OF ESSENTIAL CHARACTER
aesthetics or more general artistic studies, from which it is dis-
tinguished by the Umitation of its subject-matter to the art known
as ' formative ', and to the works which this creates. It is not con-
cerned with the psychology of the artist or spectator ; it seeks
nothing but experience from works of art, the period, provenance,
and cultural background of which are definitely known. As a
science it follows a course distinct from history, philology, psycho-
logy and aesthetics, studies which can only be regarded as sup-
plementary to it. But it stands in close touch with similar investi-
gations, based like itself upon historical data, in other spheres of
art such as poetry and music, or those fundamental departments
of life, which I have partly enumerated above. Even with these it
cannot co-operate until it is firmly established on an independent
basis.
This book has demonstrated how closely all these different
spheres of life are interconnected and how largely they determine
the fate of the formative arts. Religion itself, and particularly
Christianity, which we took as our guide, is seen on a closer
inspection to recede far into the background. All the grovv1:h and
unfolding of the first millennium and a half were made intelligible
in the first instance only by an examination of geographical and
racial factors. Despotism of Church and State over communities
came afterwards. We have seen how an attempt was made during
the fifth and sixth centuries to weld the collateral national units of
the fourth century into larger groups, and how the individual
nations of the East resisted this movement and found a support in
Persia, the second great state of the time, which, though not itself
Christian, favoured the Christians so long as they were opposed
to Rome. Thus local forms of art managed to persist, and sub-
sequently to break through the restrictions imposed by Church and
Court. This led, in the West, to the creation of ' styles ', modified
at their source by the influence of migrations and the intercourse of
peoples, and reinforced in the North by indigenous movements. In
this evolutionary process the North and South, the East and West,
proved stronger than Church and State, but differed among them-
selves in age, economic conditions, and individual character.
The preceding chapters have, I hope, made it sufficiently
clear that in early Christian church construction there were
originally three distinct manifestations of art, which first de-
veloped side by side, then formed alliance, though lacking any
close natural affinities. These manifestations were : the type of
AND APPLICATION OF COMPARATIVE METHOD 191
building ; its originally non-representational form of ornament ;
and finally, representation. The long church is imposed upon
that with transverse axis, and the dome upon both ; the church
with decorated vaults triumphs over the wooden-roofed basilica,
while representation supersedes the earlier non-representational
style, and eflFects a gradual but inevitable compromise with every
kind of structure and its decoration. I have already attempted to
show how these conflicting groups are to be defined in regard to
provenance, period, and social background ; my present purpose
is to give a comprehensive view of Christian art resulting from
their combined influence. I am assuming that Vignola's Gesii
marks the close of its career, and that the most important turning-
point in its development was the passing of the leadership from
East to West, where the fusion of elements once geographically
distinct appears to have been finally consummated about the year
1600. It is no part of the purpose of this volume to follow the
course of subsequent history. My present concern is with the
original geographical distribution of types, their later succession,
and the manner in which the ruling powers strove to subordinate all
the arts to their aims. In the first half of the seventeenth century
the forces of Court, Church, and humanistic science generally were
united, almost without distinction, in their eff"orts to suppress
every kind of individualist movement, especially in so far as
religion or art were concerned.
It was hardly possible until quite recent times to view objec-
tively the complex movements preceding the year 1600 : heirs
of that age, we were still too near it to see things in proportion.
Moreover, systematic research in the history of art has hardly yet
begun, at least not in an organized form. I hope therefore, as one
who has made a lifelong study of the subject, to meet with a
certain degree of indulgence, and to be spared instant and un-
measured rebuke for presuming in this concluding chapter to
attempt so difficult a task. It is a task to which I have addressed
myself in a practical way in all my writings, and lastly in my
Armenia ; in my belief, it is one which the specialist is bound to
undertake, and its avoidance betrays the comfortable evasiveness of
the dilettante. To make mistakes is human ; to evade difficulties
is not merely lazy, but shows lack of courage or incompetence.
Science is ill served by irresolution and resort to catchwords ;
the abuse of vague generalizations, the exclusive concentration
upon facts of space and colour, with a view to what are called
192 INVESTIGATION OF ESSENTIAL CHARACTER
objective standards, these are the subterfuges of those who will
not make the effort to think out the whole subject for themselves.
A scheme for the systematic investigation of essentials can only
grow out of actual experience ; it can be no fruit of a one-sided
or idly speculative mind.
The present book contains an examination into the origins
of Christian church art by one who has specialized in the study
of formative art. In the course of this examination certain lines
of thought have developed, foreign to those which art-history and
archaeology have been accustomed to pursue. Most important,
perhaps, is my rejection of the old view that surviving monuments
can be relied on for decisive results, either per se or as factors
determining date ; this is a view which would place history at
the mercy of chance. I hold that monuments are merely deter-
minants of their inherent values and consequently of their composite
nature to which these values contribute. Such is the method of
what I call comparative research in essential character. In this
kind of research we are concerned not with the monument itself,
but with the units of artistic values to be discerned in it ; and our
retrospective inferences relate less to the combined effect of these
values at any particular monument than to the values themselves
as units conditioning its essential character, and suggesting com-
parison with other monuments in which they may be quite
differently combined. It is necessary at the outset to formulate
some sort of arrangement as to the essential character of formative
art, the values of which it is composed, and the manner of their
investigation. This arrangement cannot be regarded as anything
more than an expedient, but its aim is purely scientific.
Formative art, like every spiritual compromise between man
and his environment, is confined to a definite vital activity, above
all to happy craftsmanship aiming at visible effects. In order to
get a clear idea of this elemental factor in art we must first study
the conditions governing the effective use of new material and
handicraft in varying kinds of physical environment. Next we
must study the spiritual values which result from man's com-
promise. In the present case it is the * mediaeval ' world which
demands our attention. We have surveyed the inhabitants of those
countries which bear upon the origins of Christian art, in the first
instance from the racial point of view. It proved that the nominal
originators of this art were the Jews, a small branch of the Semitic
race, but that the true agents of its transmission were a greater
AND APPLICATION OF COMPARATIVE METHOD 193
Semitic group, and two Aryan groups, the European and the Iranian,
of which the latter, together with its country, has been completely
neglected in the past. We then examined separately two distinct
forms of artistic creation, architecture and its decoration; finally we
divided the latter into two sections, decorative and representational,
corresponding with the divergent tendencies of the two Aryan
groups. It remains, in conclusion, to introduce into this inquiry
the order and system befitting the character of formative art, since
it is our express purpose to fill a gap which has seriously impaired
the value of all previous theories. The new point of view, which
I thus introduce, has the merit not only of independence but of
freedom from all the violence which has prejudiced hitherto the
systematic treatment of our study. Wickhoff sought to explain
its origin in West-Roman, Riegl in East-Roman imperial art ; Sybel
derived it from Hellenistic lands, Rivoira from Rome and Italy,
while Schmarsow rejected the historical in favour of an aesthetic
interpretation. But the problem is one which can only be solved
objectively and without personal bias by comparative study on
a wide geographical basis ; it is for this reason that we must begin
with a classification of artistic values. Problems such as this can-
not be solved by limiting our view to sculpture and painting, or
by putting on blinkers which prevent us from seeing anything
except representation, idealizing or realistic. I shall follow a
scheme, the value of which I have proved during the work of
a lifetime. I tabulate it here in its most condensed form :
I. Handicraft. II. Spiritual Values.
World.
1 . Material and workmanship Significance Appearance
** i
■g Objective limitation 2. Subject 3. Shape
M J
c I Personal freedom c. Content 4. Form
I. Handicraft
It is generally supposed that in a period of advanced civiliza-
tion, like that in which Christian art arose, local conditions of
material and work are a negligible factor. The best example of
this theory is that a uniform artistic taste, call it Hellenistic,
Roman, late-Roman, mediaeval or what you will, was able to
impose a uniform architectural type, the timber-roofed basilica.
245J o
194 INVESTIGATION OF ESSENTIAL CHARACTER
This is a superficial point of view and has had the deplorable
results which we have already observed. It is true that in large
towns artistic taste can be controlled, because craftsmen from all
quarters of the earth congregate there, and because authority
wields the necessary economic power. But the really decisive
influences must be sought not in the towns but in the wide pro-
vincial regions. Here art long remains the natural and unaffected
expression of humanity, servmg the interests of the community
and not the arbitrary will of autocracy ; here handicraft is
determined by local conditions. A general consideration of this
question in so far as it affects architecture will not be out of place.
I . Material and Work. These are stubborn factors, of which
no great town with all its economic resources can make itself per-
manently independent. Our proper course would be to begin
with a map showing the distribution of the indigenous building
materials available in those regions of the earth which bear upon
the origins of Christian art.
A. Building Material. Any one who makes a comparative
study of art on a geographical basis is practically driven to the
conclusion that in the vast majority of countries wood was the
original building material. Even the early history of Christian
art bears out this hypothesis. The important point is not the intro-
duction of the timber-roofed basilica, but rather the contrast
between Mesopotamia and Iran ; at that time both of these
countries were included in the Persian empire ; one of them
introduced the barrel vault into church architecture, and the other,
as Armenia proves, the dome. This contrast is surprising because
in both countries the national building material consisted of
unburned brick. The different manner in which it was employed
is explained by the fact that in Iran Aryan wooden construction
naturally led to the dome, because short beams had to be laid
across the corners of a square to support the roof, while in
Mesopotamia, on the other hand, there was no such necessity.
It is remarkable that in this significant matter Mesopotamia
has received as little attention as Iran. This is presumably because
wooden architecture was forgotten at an early date except in a few
countries such as China, where it maintains its predominance to
the present day. In India and Greece the earlier style of wood-
work was at first retained unaltered in stone, and only subsequently
underwent modification in conformity with the character of the
new material. In Iran and Mesopotamia the use of unburned
AND APPLICATION OF COMPARATIVE METHOD 195
brick prevented the introduction of stone architecture on a wide
scale ; these countries thus became, each in its own way, the
home whence the vaulted architecture of Christian times was
derived. This was the type which ultimately prevailed, and it was
for this reason that we have paid less attention to the long timber-
roofed building introduced through Christian Hellenistic in-
fluences from the Mediterranean coast. There are two reasons
why this hypothesis, of so cardinal an importance for the origin of
Christian art, has never been advanced. The first, though perhaps
not the principal, is the deplorable narrowness with which students
concentrate their gaze upon Rome and the Mediterranean. They
do not think it worth their while to search the East for traces of
Christian art, and indeed meet my pioneer work with a hostility
which is the measure of their prejudice. The second reason is that
no monuments of early Christian art have or could have survived
in some of the important regions, simply because the only building
material was unburned brick. In these regions the remains of early
churches, or at any rate of their upper structures, would long ago,
even under the most favourable conditions, have been reduced to
mere heaps of dust, though their foundations may still in some
cases be discoverable by excavation. But who has ever thought
of excavating for Christian churches in Mesopotamia or Iran ?
Unburned brick is as perishable as wood. Only in the dry atmo-
sphere of Egypt, Seistan, or Chinese Turkestan could it survive to
present times. Extant remains in rubble concrete (Palaces of
Pars), burned brick (Tak-i Kisra), or stone facing (Armenia)
are the only evidence from which we can infer the previous
existence of such buildings. Such remains suggest that there
must once have been a widespread form of national architecture
East of the Tigris and Euphrates, which has been most unfairly
ignored by students of art.
Art is conditioned at the outset by the soil and by the flora
from which it springs ; but its diff'usion depends upon a system of
centres and follows commercial currents and counter-currents,
which are, in turn, determined by political forces. Constantine
created such centres of Christian art in Rome, Constantinople,
and Jerusalem, his successors in Milan and Ravenna. It is
questionable whether these towns were situated in the line of
natural trade-routes. They are outliers ; their true area of dis-
tribution is sometimes very far away. The main lines of diff"usion
are independent of such artificial nuclei ; they are the natural lines
02
196 INVESTIGATION OF ESSENTIAL CHARACTER
of commerce, though in a lesser degree migrations, pilgrimages,
and other factors are also effective in determining their direction.
Instances of the adaptation of an alien technique to a new environ-
ment may be observed in the manus gotica in Gaul, the magistri
commacini in Upper Italy, the Crusaders with their ' Gothic ' huts
in Palestine, and ancient Persian as well as modern Italian masters
in stucco. Moreover artisans were themselves at times forcibly
transplanted, notably in the case of luxury trades. All these
are factors which at various times influenced the growth of
Christian art.
B. Work. Such were the conditions affecting material and
its manipulation, out of which the various types of work examined
in this book were evolved. These were, in architecture, the
wooden roof, the barrel vault or dome, and the column, the pier,
or the wall ; in decorative art, the organic style of wooden archi-
tecture in classical and Gothic times, or the wall-linings em-
ployed in the countries of unburned brick. Here we can only
indicate in a few words the profusion of forms from the very
outset open to Christian art, growing up as it did on the border-
land between East and West. Religion itself became the agency
through which the different indigenous styles were exchanged and
diffused. Our ideas on this subject have hitherto been far from
clear ; the extraordinarily fascinating art of the first millennium
and the subsequent development of ' styles ' in the West have
always been regarded as a purely European affair, which is
by no means the case. The court art of the ancient East, or, later,
of Rome and Byzantium, never exercised so decisive an influence
as the various national styles of architecture ; these were of wood
in the North, of varying ma:terial in the South, of brick in Meso-
potamia and Iran, accompanied, in the latter case, by an almost
unimaginable variety of decorative linings. Just as the construc-
tional basis of development is best inferred from Christian vaulted
building, so is the decorative foundation inferred in the most con-
vincing manner possible from the art of Armenia and Islam.
Historians have hitherto regarded Christianity and Islam, the
last two religions with historical founders, as embodying two
diametrically opposed tendencies in art. That they might both
have sprung, to some extent, from a common source was suspected
by nobody, except in so far as the traditional reference was made
to classical influence. The development of Christian art was
determined by the fact that it chose the interior of buildings for
AND APPLICATION OF COMPARATIVE METHOD 197
decoration, and that it lived out its allotted time in the Mediterra-
nean area ; Islam chose the open court, and in the field of decoration
entered upon the inheritance of Iran. In Europe, curiously
enough, the ancient timbered style was again revived for roof con-
struction under Southern influence at the end of the Roman
period. Of the purely wooden architecture characteristic of the
Northern peoples nothing has survived from the first millennium,
and its artistic qualities are practically unknown. But Scandi-
navian wooden churches of more recent date plainly indicate
whence ' Gothic ' derived much of its organic character ; the
Oseberg ship. Urn as and early Teutonic antiquities suggest that
the birth of ' Gothic ' did not so much imply a complete change
arising out of stone construction as the logical development of the
feeling for organic growth. Its original non-representational style
was sacrificed to an increasing luxuriance of figures. Italy
borrowed the idea from the North, taking the figures out of their
architectural setting and adapting them to a nearer view. I shall
not refer to painting again at this point.
The introduction of the decorative wall-linings of the East
into the Roman basilica, and particularly into the vaulted apse,
led to a completely new orientation in the art of the Western
Mediterranean as well as in that which succeeded it in Europe.
The Eastern ideas which accompanied vaulting were incor-
porated in the wooden-roofed basilica ; we thus have the most
surprising of alliances, that between an effete organic architecture
and decorative wall-lining. The combination at a later time found
approval in the North, where in wooden buildings decoration
is confined to the covering of surfaces with design.
Whether I set out from East or from North, I invariably find
confirmation of the conclusion reached by Gottfried Semper,
though by a wrong use of historical evidence, that material and
work are prime factors in the development of art. And again,
if I consider that modern as distinct from classical art in Europe
begins with the period of the great migrations, I still find
it necessary to place the influence of craftsmanship at the head,
whether it finds expression in plaited or woven materials, or in
carved or incised work in wood, metal, or leather. At the
beginning of the Christian era the Germanic peoples pressing
down towards the South still maintained that intimate feeling
for craftsmanship, which goes back, apparently, to neolithic times,
and served as a foundation for their whole sense of artistic
198 INVESTIGATION OF ESSENTIAL CHARACTER
values. The need of the subject or the figure was rarely felt by
them ; they were interested only in things and their decoration.
II. Spiritual Values
The religious man sees the world through spectacles, only
removable when he is free from the dictates of the Church, or
of State and Church acting together. Both of these forces
victimize ; artists, nations, and cultures are all sacrificed to
the lust after expansion of power. We have observed how com-
munities originally created their own art, and how at a later time
Church and State intervened with fatal effect. Freedom is
essential to the artist ; his individuality shrinks at the tyrannical
infliction of restraint ; serving the artistic ends of power, which
makes him its tool, he degenerates only too often into the
journeyman of art. The result is seen in a monotonous repetition
of subjects and figures ; form loses its content, and development
runs on to the worn-out rails of * style '. It is the task of research
to reveal a flowering-time of art only possible through religious
freedom and often only attained at the price of bitter conflict ;
it was a time so fruitful that it may inspire us even to-day. On the
other hand so-called * styles ' may be of service to modern crafts-
men, and in some degree, though less directly, to the artist.
Research clears vision only when it interprets, not when it merely
classifies and describes ; only so can it reveal the origin of deter-
minant values. Thoughtful artists even more than the educated
public of the day have the right to demand such interpretation
from research.
The general historian may find his account in giving special
prominence to communities of artificial origin, in this particular case
Church and State ; but for the specialist, soil, nationality, and the
individual artist are the obvious foundations of research. Personality,
in the sense of a natural unit, whether of the mass or the individual,
is the only creative power, and not the artificial community. A
history of art which attributes greater importance to the exploiting
will than to original creative impulse may comprehend a fettered
imitation, an art which is no art ; it is helpless to explain that
which casts into the stream of development new and fruitful ideas.
The recurrence of the same artistic values throughout Europe in
the sixteenth and seventeenth centuries is due partly to the
infectious influence of fashion, which accompanied the north-
AND APPLICATION OF COMPARATIVE METHOD 199
ward march of Court, Church, and Humanism, and was only
resisted by a Rembrandt at the cost of poverty ; partly also to
a counteracting assertion of Northern influence in the South.
In early Christian times the growing power and wealth of Church
and State produced the same results.
Religious movements have rarely confined themselves to the
lands of their origin and growth ; ecumenical religions, in par-
ticular, spread their conquests far and wide. Christianity did not
expand at a stroke like Islam ; it knew centuries of freedom.
It was not until about a.d. 300 that it became a militant power
spreading in every direction. But it was during the earlier
centuries of freedom that those forms of art arose which were to
determine the course of its later development. We must not be
misled by the fact that Rome in the fourth and fifth centuries,
Byzantium in the fifth and sixth, each found its own direction
within the compass of the wider currents. It was not the arbitrary
welding of artistic values in these places which counted, but the
units which were welded, the values which, in despite of Church
and State, brought their native quality in all its freedom to peoples
always, or at least for periods, unswayed either by Byzantium or
Rome.
Significance. The opinion has been held that the essential
character of Christian art may be inferred from the deeper
meanings by which it is penetrated. This view involves contra-
dictions, since there is obviously present a simultaneous and very
powerful influence acting immediately upon the senses. One need
only call to mind the recognition, not to say the admiration,
bestowed upon mosaics by contemporary art critics ; in this result
the chief part is played by mere colour ; remoteness from nature
comes second, seeming pure expression. The juxtaposition
presents a problem which can only be solved if we remember that
Christian art combines within itself a number of distinct artistic
currents, not to be confused, and springing from the most diverse
sources, historical, geographical, racial, national. The gradual
discovery of the East was a necessary preliminary to its solution.
Byzantium was the first to be revealed ; then came Asia Minor,
Syria, and Egypt ; finally, Mesopotamia, Armenia, and Iran;
for me, it is now the turn of Northern Europe — so many names,
so many distinct streams of influence, or transformations due to
the blending in varying measure of such currents. These complex
values are very different from those of Mazdaism or Hellenism
200 INVESTIGATION OF ESSENTIAL CHARACTER
which preceded, or those of Islam which succeeded them. But
taken smgly, they yet have their points of resemblance. It would
be easier to define the character of each of these religious units by
its negative than by its positive qualities. But the crucial factor
in Christian art is not, as might appear, an arbitrary and enslaving
will, but rather the vital impulse which reached it from the East
and inspired it to fresh creative effort. This is especially apparent
under various forms in the transitional period, when it passed from
the East and South to Europe and the North.
2. Subject (Purpose). The form of the church was not at
first everywhere or exclusively determined by its function as a
place of assembly for the community. In Armenia, for example,
that is to say in the ancient Mazdean cultiireaTCarit appears to have
been permanently influenced by the form of the founder's tomb,
although nothing is actually known of the tomb of Zarathustra.
The same influence, albeit in an indirect and borrowed form, was
fossibly stronger even in Byzantium than is at present recognized,
n Nisibis a church and a tomb-like baptistery still stand side
by side in their original form except for their roofs, which have
been partially restored ; the baptistery, together with the porch,
is built on the South side of the church. The kind of building
which followed the death of Saint James (a.d. 338) does not
seem to have become the general rule. InJ^imenia the actual
founder's tomb was enlarged for use as an assembly hall, the
, single building performing a double function, like the church of
'^ S. Peter in later times. In the fifth century, when the Church
wished to introduce the vaulted long building as the type, it found
an obstacle in this firmly-rooted style ; the result was the fusion
of the domed with the long-naved building.
It was otherwise in the Mediterranean area. There the
timber-roofed long building was the prevailing form of assembly
hall from the earliest times. In Jerusalem, Constantine placed it
next to the tomb, from which it was separated by an atrium on
an axis running from West to East. It would be a task in itself to
make a thorough study of the relations between these two practical
types ; but we may note at once that at the tomb of S. Vitale in
Ravenna, as well as in that of Charles the Great at Aix-la-Chapelle,
the domed building served the second purpose of an assembly
hall ; on the other hand, the tombs of Constantine in Con-
stantinople and of Galla Placidia were domed structures, built
from the first adjoining a cruciform church. Is not this a clear
'?
•^e
AND APPLICATION OF COMPARATIVE METHOD 201
case of the partial influence of Eastern usage in the relationship
between the form and function of a building ? This is a question
which naturally does not arise in the case of the timber-roofed
basilica of the West. The building with a main axis existed there
from the very first ; but it is foreign to the church of Armenia in
its earliest period. In that country the central square bay has
four apsidal buttresses, of which that on the Eastern side appears
by an afterthought to have been converted into an apse when
a dais for the altar was introduced. Discoveries in Carinthia also
indicate that the altar space and the hemicycle where the clergy
sat were probably quite distinct in origin.
The fact is that the method of orientating churches varies in
the West and on the Eastern coast of the Mediterranean, whereas
he apsidaLhuttr^ses of Armenian domed churches without
exception mark the four cardinal points. Here, as well as in Asia
Minor, the long church was consistently orientated towards the
East, and from these regions the practice was diffused in all
directions. On the coast of Syria, and in Egypt and North Africa,
the introduction of an Eastern apse opposite to the ancient
Western apse produced the church with double choir, a type of
which we seem to trace the influence on the early art of Western
Europe. It is not impossible that the practice of orientation in
the pioneer districts of Armenia was due to pre-Christian ideas,
connected perhaps with the Mazdean sun-cult ; in this connexion
we may recall the association of dawn clouds with the ideas of
redemption and of the Last Judgement (p. 180).
On the other hand the later combination of church and
assembly hall appears to have been due to the influence of classical
temples, and to occur first in the church in which the main axis
is longitudinal. My attitude is, throughout, that of an observer ;
discerning theologians will find suggestion here without need of any
lessons from me. Was the whole of the church treated as a place of
worship, or was only a certain section of the large assembly hall set
aside for this purpose ? In Armenia an altar-dais was introduced
as a special feature in the Eastern portion of a building radially
planned round a central bay with a high dome with vertical axis ;
this altar was raised about three feet above the space left for the
congregation, and approached by two lateral flights of steps. The
unity of the building was thereby destroyed. One seems to see
the origin of those later Hturgical representations of the Last
Supper, in which Christ appears as priest by the ciborium
202 INVESTIGATION OF ESSENTIAL CHARACTER
distributing bread on one side and wine on the other to the apostles
approaching from right and left.
In the Mediterranean area the altar was placed at the end
of the long nave. This arrangement had the merit of preserving
artistic unity ; but it introduced a perpetual conflict between the
upward view into the dome (even after its removal from its
original central position) and the perspective view into the apse.
In the Mediterranean the only type accepted by the Church
was the long assembly hall with the main axis longitudinal ;
where we find the domed type of building, we must ascribe it to
the interference of temporal power, which introduced it from the
East in order to gratify its desire for effect. This is the most
probable explanation of S. Lorenzo at Milan, and certainly of
S. Vitale at Ravenna and the imperial chapel at Aix-la-Chapelle.
The plan of the two latter buildings was not determined by the
founder's tomb alone. The introduction of galleries suggests
that they were also intended to serve as imperial assembly-halls.
It was otherwise in Armenia ; there the dome originally prevailed
to the exclusion of all other types ; Zwarthnotz was probably also
intended by Nerses III to serve as a kind of imperial hall. But
after the alliance of the Armenian Church, in the fifth century,
with Nisibis,Edessa,and Byzantium, a tendency towards length of
nave set in, and this type for a time prevailed, in accordance with
Canon 182, though still in conjunction with the dome. It thus
appears that the tendencies of East and West were in opposite
directions. A compromise was first reached in the intermediate
region, the dome being generally adopted bythe Orthodox Church.
In the West the influence of Iran secured its first triumph during
the height of the Renaissance in the church of S. Peter, its second
in the church in which Vignola crowned the attempts of Leonardo
and Bramante by constructing his famous domed hall-church for
the Jesuit order. In just such a way Armenian architects had
succeeded centuries before.
A. Non-representational Decoration. The decoration, like the
construction of churches, had originally a practical aim ; it had
significance without any second purpose alien to the principle of
art for its own sake. Its object was to heighten the effect of the
exterior mass and the interior space of the building, and this was
achieved by the use of light and colour, the starting-point being
line and ornamented surface ; the human figure was excluded,
as in the case of the nomadic and Northern races. This type of
AND APPLICATION OF COMPARATIVE METHOD 203
decoration seems to have been permeated with symbolism even
in Mazdean times. Animal- and plant-motives, introduced as
vehicles of Hvarenah, that is, the glory of God, imparted a religious
atmosphere to the purely decorative wall-linings. This style
was transmitted both to Christianity and Islam ; the former
developed its symbolism still further and combined it with repre-
sentations of a purely objective character.
Sculpture and painting are usually called the representational
arts in the narrower sense of the term ; but this is incorrect. There
are many artistic provinces which practised sculpture and painting,
but did not represent ; the religious art of ancient Iran was found
to belong to this order, and also ancient Chinese and Teutonic
art. Again, following the track of the nomadic races south-west
of Iran, we came upon the same thing among the Jews and Arabs.
Christian art, in its earliest period, belonged to this non-representa-
tional group of Northern races and pastoral nomads, who used
ornament of wood, stucco, and stone, and painted in colours, but
did not represent. I was compelled to take architecture as my
starting-point, since Christian art did not originally, like late-
Hellenistic, produce paintings or sculptures designed for near and
intimate contemplation, but confined itself to the decoration of
buildings and objects of personal use. When figures of Christ and
the chief apostles were first produced in the fourth century, they
caused general indignation even among the Greeks and Ara-
maeans, who later became the most determined agents in the
diffusion of picture-worship ; so great was the change wrought
upon the spirit of Christianity in the course of time by Church,
State, and general culture. Hellenism, with the patronage of
Church and State, triumphed over popular non-representational
art ; the latter revived somewhat under the influence of the
Middle Ages and the iconoclastic movement ; the Renaissance
once more accepted the sole dictatorship of Hellenism, while
* Baroque ' made extravagant use of both styles. We are only just
beginning to recognize the primordial nature of non-representa-
tional art, and many years must elapse before we can fully realize
its fundamental significance.
The decoration of Christian churches was designed for mass
effect and intended to be seen from a distance. The first Christian
masters had no thought of figure sculpture such as that of Greek
temple architecture, nor of painted representations like those
found in the interiors of Pompeii or in the Catacombs. The nature
204 INVESTIGATION OF ESSENTIAL CHARACTER
of the wall was not destroyed ; it retained its character as a flat
surface covered with a decorative lining. It is therefore futile to
confine the treatment of early Christian art to a study of types,
and this study itself to the region of sepulchral art. This is
deliberately to ignore the crucial importance of church art, and
to substitute false light for the true.
B. Representational Decoration. Like architecture and the
decorative art associated with it, representational decoration
differs in character according to its origin. The differences are
in this case, however, not geographically conditioned, but by the
will of authority and by racial factors. The East- Aryans and
Armenians drop out, except for the fertilizing effect of Mazdean
ideas which were translated by Southern art into representational
form. Representational art derived its distinctive features from the
West- Aryans and from the Semitic empires. The West- Aryans had
a preference for symbolic representation until Greece and Rome
adopted the policy of the old Semitic monarchies . The chief vehicle
of tnis symbolism, as in all representational art, was the human
figure. Originally the West- Aryans never depicted God as a human
being endowed with power ; this was an illusion practised by the
Semites, though not by the Jews. Indian art, again, rigidly excluded
the anthropomorphic treatment of God until the Hellenistic style
was introduced into Gandhara.
In place of architectural decoration, designed purely for
artistic and symbolic effect, the Church about a.d. 400 introduced
representation for didactic purposes, and with it Semitic traditions.
Previous to this date the non-representational style of Iran and
Armenia, as well as of Judaism, had prevailed, at least so far as
church architecture was concerned. As this representational
art flourished on the same soil from which the Buddhist art of
Gandhara derived, it probably grew up at a period during which
intercourse between Asia Minor and India still existed In
monumental art, however, symbolic representations of the idea of
redemption had established themselves from the very first, as
I have already explained. It is significant that no trace of this art
was introduced into Church decoration — Achthamar, a.d. 915-921
is an exception — and that its development shows a sudden and
abrupt transition from the decorative to the didactic style.
Generally speaking, the Semitic style very soon began to
predominate in church representation ; this style was essentially
realistic. Thus Christ appears as a Semite, generally as a judge
AND APPLICATION OF COMPARATIVE METHOD 205
enthroned with the Christian book of the law in his left hand, and
his right hand raised ; the Virgin appears not as the simple
Mother, or as interceding for mankind, but enthroned as the
Mother of God . Both are surrounded by figures of angels and saints ,
who keep the Founder at a proper distance, and thus resemble the
figures of the minor deities in a Gudea relief at Berlin. The hope
of redemption is no longer depicted in parallel scenes from the
Old and New Testaments, but is associated with the Apocalyptic
conception of the Last Judgement. This is supplemented by
didactic biblical scenes resembling in treatment those Semitic
series of biographical subjects dictated by the ruler's insatiable
desire to dazzle the people by personal display and to give weight
and permanence even to the most trivial details of his life. Self-
glorification and irresponsible autocracy determine in like manner
these representations of the deity.
The stream of Semitic representation appears to have sprung
from the intellectual centre in Northern Mesopotamia, though
probably not before the severance of the old Hellenistic relations
which had connected Asia Minor with India ; the chief gate through
which it communicated with the Mediterranean was Antioch, the
secondary gates to North and South were Constantinople and
Jerusalem. This was the style of church art which won permanent
acceptance ; one of its aims was to depict God, or the Church, as
the Judge and Teacher, another to instruct the illiterate, a third to
make such dogmatic manifestoes as that on the triumphal arch
of S. Maria Maggiore or in the choir of S. Vitale. It maintained
these aims simultaneously throughout all periods. Artistically
speaking it hardly ever rises above the subservient role assigned to
it by the Church, and the artists were, naturally, never mentioned
by narne. Subsequently, in the sixth century, came the intrusive
influence of the Byzantine Court. The cycles fixed by the Church
were heightened to the point of dazzling effect by accentuating
magnificence, and, in the representation of persons, by displaying
Christ and Mary in all the bravery of court splendour. The
climax in effect was reached when angel guards take the place of
human witnesses. A decisive impetus seems to have been given
to the cult of the Virgin by the imperial foundations at the Church
of the Nativity at Bethlehem.
In the West the ideas transmitted from Nisibis by Cassio-
dorus flourished and grew up into a great system which clearly
betrays the didactic basis, and the nature of that ecclesiastical
2o6 INVESTIGATION OF ESSENTIAL CHARACTER
philosophy which was only broken, in so far as it affected art, by
the rise of lay craftsmen in the towns.
The idea of the subject as now understood was first intro-
duced into Christian art by the theological schools of Nisibis ;
the idea, that is, of an event depicted for its own sake, and so in
the present case, historically. Here we see the old Semitic
spirit at work. The conception is inherent in a didactic and
hieroglyphic style ; the one aim is the clear rendering of the object.
It is well illustrated in Byzantine art and, earlier, in the school of
Gandhara.
This ecclesiastical influence very soon turned to dogmatic
illustration. Learned expositions of dogma were evolved like
those on the triumphal arch of S. Maria Maggiore, in the church
of S. Vitale, and other places ; it is a tendency which permeates
the whole of mediaeval art, attaining its due consummation in
scholasticism and the works of Thomas Aquinas. Mere objective
fact is emphasized to such a degree that all other elements recede
into the background and form is, as far as possible, replaced by
geometrical division. The student cannot be too clearly warned
not to regard this changed view of the significant as the main
factor in the development of mediaeval art, as has recently been
done in a work entitled ' Idealism and Materialism in Gothic
Sculpture and Painting '.
Appearance. It is true, as a general observation, that
religious emotion creates forms, and that the Church stereotypes
these into permanent shapes. Hence it comes about that only the
first four centuries of our era were essentially creative. In this
period arose, or were selected from the previously existing stock,
those appearance -values which give to early Christian art its
essential character and supply the materials upon which later
artists worked, bringing them together in various ways to form
fresh units, or subordinating them to new values conditioned by
place, time, and cultural environment. In the West the move-
ment began with the traditions of the Mediterranean area, those of
the Semites and the East- Aryans. The great migrations led to fresh
developments in this region, and to the adoption of new types
differing from those of early Christian times. A logical develop-
ment of Eastern values took place in the West, leading eventually
in the North to the so-called ' Gothic ' style ; this North-Aryan
creative achievement was unhappily brought to a premature close
by the ill-considered adoption of Italian Renaissance art. A few
AND APPLICATION OF COMPARATIVE METHOD 207
great masters now stood out in intellectual prominence ; but the
general masses merely accommodated themselves to the demands
of State, Church, and Humanism ; in this way there originated
styles which spread throughout the whole of Western Europe.
Eastern Europe, for the most part, remained faithful to the style
established by the Orthodox Church.
The units of shape and form fall into two distinct groups :
the first of these is characterized by its imitation of nature or of
types derived from earlier art ; the second by its independent
treatment of mass, space, light, and colour. We might say that
the whole of Christian art, from its beginnings to a.d. 1600, lies
between a world of nature-imitation in process of decay and
another world in which that imitation was deliberately resumed.
Had not the spirit of Nisibis won the day with its objective method
of edification, carried into effect by Syro-Hellenistic types, the
effort to attain form through the symbol might have remained the
fundamental principle of Christian art. The Iranian East and
the Teutonic North both, like the Hellas of earlier times, strove
in this direction, though the means which they employed were
different.
3. Shape. Christian Church architecture sprang from such
diverse sources that we cannot say of any particular shape
that it was typical, in the sense in which the classical temple
was a type. This diversity embraced not only raw material and
technique ; it extended also to the use of the building whether
as founder's tomb, place of worship, baptistery, or assembly hall,
as the case might be ; thus the artist's freedom was necessarily
limited by the conditions under which he worked, even when
authority permitted him to exercise his own choice in the matter
of form, provided always that he produced something impressive
and monumental in character. When he worked in the central
seats of Church or State he adopted such outward forms as he
found ready to hand. He created no new types of building, but
merely heightened the effect by increasing the proportions of those
already existing. In the matter of decoration he was equally un-
inventive, contenting himself with introducing Eastern features
into the West, just as, at an earlier date. Western features had been
brought to the East through the medium of Hellenism ; such was
the state of affairs until the North asserted its influence, to be
followed almost immediately by the humanistic and classical
revival in the South.
2o8 INVESTIGATION OF ESSENTIAL CHARACTER
A. Constructional Form. Historians of art have always
shown a predilection for this in their works ; they have classified it
into * styles ', laying special stress upon similarities in ground-
plan, elevation, and ornament. In the sphere of early Christian
architecture this was simple enough. The basilica was taken as
the type ; it consisted of a central nave between columned
arcades and terminating in an apse ; on either side was an aisle
closed on the outer side by a wall ; above was a clerestory.
Domed buildings did not fit conveniently into this scheme ; but
as they only occurred in isolated instances, they could be treated
as belated survivals of classical architecture.
My books Kleinasien and .4m?V/a, published in 1903 and 1910,
made us acquainted with the vaulted long churches in Cappadocia
and Mesopotamia. The facts were disturbing, and every effort
was made to force the date of these churches mto the mediaeval
period. There was much excitement in certain quarters ; accusa-
tions of dilettantism were made, and even of disingenuous pro-
cedure. What will be said now that Iran and Armenia have come
into the very foremost rank with their domed architecture ?
An end has been made of the easy old classification by ' styles ',
each with its distinguishing marks, counted on the fingers, and of
its modern substitute, faith in the sole dominion of arbitrary laws
in the historical explanation of style. Three architectural types,
each with its own wide area of distribution, existed side by side as
early as the fourth century; these were : (i) the East- Aryan dome
upon a square plan ; (2) the Semitic church with transverse nave,
vaulted in Mesopotamia, and furnished in East Syria with the
South-Arabian stone-roofed arches; (3) the West-Aryan timber-
roofed long building which was given a vaulted roof in the East.
These three types maintained a separate existence, despite the
r aggression of the West- Aryan church. The dome soon triumphed
in the middle region at Constantinople ; the barrel- vault at a
later period in the West, until here too the dome at last made its
way with the Renaissance. These types were not Hellenistic in
origin. Only the basilica was Greek in its three-aisled form and its
e^ ^ f clerestory ; its timbered roof was displaced by forms indigenous
y >r ^ to the East.
Q^ '_Y Amid this surprising diversity of local forms which became
^ Y- t'^ distinctive types, it was not so easy for the Church entirely to
■^ check the diffusion of national forms by the different peoples.
But accepting, as she did, only forms sanctioned by liturgical needs,
AND APPLICATION OF COMPARATIVE METHOD 209
she in fact succeeded in imposing the long-naved type of church.
At Rome her success was complete, not so at the imperial capitals
of Milan, Ravenna, and above all Constantinople. These seats of
temporal power were, indeed, quick to recognize the single dome
as a form of architectural expression ; its demand for the absolute
subordination of all the parts to the whole agreed with their own
principles, and they set it in effective opposition to the demands
of the Church. While Roman church architecture was stagnating
in the monotonous uniformity of the timber-roofed basilica,
Armenian and other architectural forms forced an entry into the
West, at first by way of the Eastern and Western imperial capitals,
afterwards by routes that left them on one side. This partly '
explains the abrupt appearance of the single dome in the West.
It was otherwise with the barrel-vaulted long-naved church, the
diffusion of which from East to West was effected by gradual
stages. The Roman Catholic, unlike the Byzantine Orthodox
Church, was never in complete subjection to a Court ; it was
thus able to hold out against the dome, though eventually
succumbing to the barrel- vault,
B. Decoration. Much is explained by the fact that, in the
matter of decoration, the long-naved church of the Mediterranean
contented itself with marble columns and painted and gilded
wooden ceiling, a limitation also observed by Constantine in the
Church of the Holy Sepulchre at Jerusalem, All other kinds of
decoration were later additions, and derived from the Eastern art
of covering walls with sumptuous linings. As regards exterior
decoration, a distinction was observed between the wooden-
roofed churches which rejected it and the vaulted church which
practised it from the very first. In the West, in the same manner,
exterior decoration was exceptional until after the introduction
of the vault.
In spite of the tenacity with which the church in the Mediter-
ranean area clung to Hellenistic traditions in architecture. Eastern
influence ultimately prevailed ; this is perhaps most clearly
proved by the fact that it was not construction but the covering
of walls with decoration which really influenced the builders.
Northern (Gothic) art, with its traditional feeling for construc-
tion derived from its work in wood, was the first to reintro-
duce organic methods, though it started from the pier and the
vault, not from the classical column. Down to Gothic times all
Christian architecture, even the columned basilica, was really so
3451 p
210 INVESTIGATION OF ESSENTIAL CHARACTER
much walling decoratively lined. Whether it was executed in the
simple and popular style, as in Armenia, or received, as at Con-
stantinople, the rich embellishment characteristic of work done
in the sphere of Court influence makes no essential diff^erence.
The change which took place in the capital is significant.
The appearance of the acanthus in Greece and of other plant forms
in the North marks the transition from pure construction to the
imitation of natural forms. Such a development was impossible
in the East, since the ideas of the simply articulated building and of
realistic representation were entirely foreign to the Eastern mind.
When in the Mediterranean area it actually came to making the
decorated capital look like a basket, and by a still more audacious
step, like a basket with beasts and birds upon its top, what we
have before us is nothing more nor less than a frankly Hellenistic
version of an Iranian idea.
The Church concerned herself exclusively with distinctive
types, preferably with those upon which it has set its official seal ;
pure form was to her of minor importance. Every movement in
the direction of artistic freedom was hampered, by regulations,
by old tradition, and by the rules in Painters' Manuals ; we must
therefore draw a clear line of demarcation between ecclesiastical
and real religious art. At the present time we take it absolutely
for granted that religious feeling must express itself by means of
the human figure. This is one of the chief causes which have
almost deprived us of the capacity to distinguish between art and
representation.
C. Representation. In the treatment of the human figure,
the essential difference between Hellenistic and early Christian art
lies in the fact that the representation of the one is taken direct
from nature, that of the other is derived from ideas. The most
impossible theories have been advanced in explanation of this fact.
I need only mention the latest, according to which Neo-Platonism
and, in formative art, illusionism are supposed to have substituted
for the sensualism of classical art a spiritualism, regarded as the
basis from which the structure of all mediaeval art was to rise. But
quite apart from the connexion of Neo-Platonism with the East,
through the Indian doctrine of Yoga (Conrady), the development
of art in the transitional period from classical to mediaeval was
not determined by the fashions and catch-phrases of large towns,
as development is to-day. The creative energies of mankind,
even within the limits of the then civilized world, were by no
AND APPLICATION OF COMPARATIVE METHOD 211
means exhausted. Later classical art indeed, whether ' immanent '
or * autonomous ', had indeed lost all creative power ; it lay in its
last throes at the mercy of the East-Aryan reaction which now
broke out after centuries of repression, and succeeded in obtaining
full recognition even in an unfamiliar domain, that of representa-
tional art. To East- Aryan influence was due that non-natural
envisagement of figure subjects which was destined to prevail
throughout the Middle Ages ; the new kind of vision implied an
earlier state of non-representational art in which abstract form
predominated. This change did not affect the catacomb paintings
or the sarcophagus sculptures, the figures of which are in the direct
line of classical tradition. It showed itself most markedly in figures
on the monumental scale and, above all, in that of the vaulted apse,
on which the eyes of the congregation were naturally focussed.
Of course no anatomist would ever think of using the figures of
early Christian art for the study of human anatomy ; even the
mosaics of outstanding artistic excellence are synthetic composi-
tions pieced together out of classical elements. This applies
especially to the drapery, to which heads and hands are merely
attached in their proper places, without any suggestion of a
lifelike figure supporting them and giving them cohesion ; much
less of that art, in which the Greek excelled, of bringing out the
significance of limb and joint through the drapery that overlies
them. The human figure was reduced to a symbol, and was used,
in the scheme adopted by the church, just like the written letter.
The figure-compositions, like the landscapes, were merely col-
lections of such symbols. The types were used to represent Christ,
the Virgin Mary, and the Apostles, and were absolutely fixed ;
they were as obvious and as familiar as the letters of the alphabet.
At this period art was even less identified than at other times
with the treatment of the solid figure. The type, such as it was,
was taken over from classical art, and only given naturahstic traits
when portraiture was attempted. The treatment of landscape has
been fully discussed in connexion with the subject of Hvarenah.
Thus the interaction of Hellenistic, Iranian, and Semitic
influences produced a remarkable change in the classical treatment
of solid figures. The three-dimensional rounding off of the body
disappears, the figure is flattened, modelling persists only in a
perfunctory way, lighting and shading to indicate depth are
abandoned or survive in the most rudimentary form. Things con-
tinued after this fashion until the great period of North-Christian
p 2
212 INVESTIGATION OF ESSENTIAL CHARACTER
art revived a feeling for organic structure. How are we to inter-
pret the change which took place at the end of the classical period ?
Late classical art itself shows a very striking development
from modelling in tone to the abrupt contrast of light and shadow,
producing a similar effect to that of contrasting colour. The
result was obtained by the use of the drill, the chosen instrument
for the purpose. The process was marked at the beginning by
virtuosity in the detachment of the figure from its background.
The impulse to this change probably came, as in India, from the
purely decorative art of wall covering in Northern Iran. The
movement very soon produced a reaction ; the surface-filling flat
figure completed by colour came into its rights ; the classical types
appeared flattened, and all sense of three-dimensional space was
eliminated. Only a meagre remnant of the antique style lingered
on in the almost hieroglyphic representational art affected by the
Semites, of which ancient models might still be seen in the old Meso-
potamian palaces and reliefs on triumphal columns . The theological
school of Nisibis then created the artistic ' Canon ' which was to
dominate the whole of mediaeval art. This still admitted some
tendency towards realism especially in portraiture, as is evident from
the ' Edessene ' types of Christ, S. Peter, S. Paul, and others. The
appearance of these types must be later than that of the Hellenistic
types of Christ and Buddha created in Asia Minor and Gandhara.
By the side of this broad stream of Semitic art certain
classical reminiscences continued to survive here and there, both
in the technique of painting and in the types presented.
The reaction from a conventional to a naturalistic art is first
seen in the architectural church sculpture of the North, which
introduced not only animal and plant figures into the ornament of
its constructional members but also those of men, and so trans-
mitted them to Italy. There the Humanistic movement, involving
the study of classical forms, and, even more, the awakening demand
for scientific knowledge, led to a predominance of observation
at close range, and, as a result, to a more or less exact imitation of
natural forms. Hence arose the common belief that correctness in
delineation was the supreme object of art. The subjective treat-
ment of form was at times almost forgotten.
4. Design. It is in the province of form that the contrast
appears most marked between the Hellenistic and Christian
classical art on the one hand, with their partly non-artistic methods,
and the new East-Aryan stream on the other. In the catacombs
AND APPLICATION OF COMPARATIVE METHOD 213
the walls are covered with representations with a lack of feeling for
interior spatial effect as conspicuous as that shown by Buddhism.
Only the Iranian East continued to treat interiors as a medium
of artistic expression in themselves and dispensed with representa-
tion entirely. This comes out most clearly in the domed churches
of Armenia, the nobility of which, like that of the earliest Doric
and later Northern (early Gothic) buildings, lay in the purity of
their architecture ; all decoration in the form of painting and
sculpture was at first almost wholly rejected, and subsequently
treated as altogether subordinate to the architectonic plan. In
Armenian buildings exterior effect was produced by mass, the
interior effect by space ; the decoration in both cases was held
in restraint.
It is otherwise with Christian classical architecture in the
Mediterranean area. There the exteriors are neglected, the
interiors form an incongruous patchwork made up to suit the
exigencies of the moment. The columns are brought together any-
how from older buildings, except in cases where the hand of
authority intervened to produce uniformity. The interior decora-
tions, executed with an especial object in Rome, Ravenna, Parenzo,
and other places, are all borrowed from East-Aryan vaulted archi-
tecture and thence transferred to the long building now regularly
demanded by the Church. This is the case with all the new forms
of the capital brought from the quarries of the sea of Marmora, and
to a large extent also with the surface linings of floors and walls,
especially those in mosaic. Owing to the way in which they were
constructed out of small cubes designed for curved surfaces,
mosaics can only have reached their highest level of artistic
excellence in the vaulted architecture of the East ; how high that
level was can be seen from the earliest and best-known churches
that have survived, so far as their mosaics are not mere imitations
of paintings or even of miniatures transferred to the available
space on wall and roof. A good criterion is afforded by the
decoration of the apse. This was at first purely formal and non-
representational in style, or consisted of a landscape with the cross;
it is only about the beginning of the fifth century that human
figures make their appearance which, in their disposition of mass,
space, light, and colour, rival in importance the structure itself
and its decoration.
A. Mass. The idea, derived from Graeco-Roman and
Italian art, that buildings should have one particular side, the
214 INVESTIGATION OF ESSENTIAL CHARACTER
facade, designed to impress the spectator, has diverted the
attention of art-historians from mass as such, from one of the
essential values in formative art. If we look at an Armenian
church, even one built after the combination of the long nave and
Lthe dome (Fig. 12, facing p. 72), we notice that it has no true
facade, but is intended to impress the spectator by the totality
of its structural mass from whatever side it is viewed. In the
timber-roofed basilica on the other hand, as in the ancient temple,
special importance is given to the gabled end. Thus in our
impression of the exterior a single wall takes the place of the whole
structure. Hildehrnnd^s Problem derForm is a good illustration of
the extent to which we have accustomed ourselves to the classical
point of view. Naively assuming the self-evident nature of his
views, an assumption only to be explained by a long and one-sided
devotion to the study of classical and Renaissance art, he treats the
question of relief as alone admissible in artistic appreciation, though
. it may be admitted that he is primarily thinking of sculpture. But
^\ A if we recall the Armenian centralizedchurch witLjiiche-buttresses
as It existed in the fourth century before the introduction of a long
axis, a building not unlike the dome of the cathedral at Florence as
it would appear without its nave, we must admit that this Christian
type of structure was well fitted to introduce a wholly new ideal,
A opposed to that inspiring the antique temple, an ideal in which
Pt^fZ^ centraliaatioRand height, rather than concentration upon the fa9ade,
were the outstanding features . Only the influence of Hellenism and
the Church, with their adherence to the classical plan with long
major axis, prevented the development of the centralized church,
which would otherwise have become the standard type in Northern
(Gothic) architecture. The remarkable contest waged aftef~the
time of Leonardo and Bramante by certain of their successors,
who favoured the execution of the original design for S. Peter's,
had been anticipated in the East more than a thousand years
earlier. The feeling for mass in architectural design clearly
exerted an influence upon Eastern painting of the early Christian
period.
The use of ornament to accentuate the effect of mass was at
first naturally confined to the East. In Armenia blind arcading
like an embroidery in delicate low relief after the Persian style
forms a strong contrast with the massiveness of the whole building ;
in Mesopotamia the edges of the roofs with their alternating bands
of light and shade, in Syria the mouldings of similar character
AND APPLICATION OF COMPARATIVE METHOD 215
round the windows, and, most important of all, the development of
a regular church fafade, derived in part from the Hittite Khilani,
all served this common end. The last-named feature was the
prototype of that later Western type of fa9ade which consisted of
a porch flanked by towers. On the other hand, door and portal in
Christian architecture were derived from Armenia, and combined
with Syrian traditions to form those magnificent Northern fa9ades,
which the Renaissance was unable to understand, because it
misunderstood the character of antique columns, and utilized them
not as working members but as mere decorative features.
In representational art the Semitic arrangement in zones
prevents the fusion into a single mass-unit of different groups,
which are sometimes massive m themselves. Only in the apse,
as once in the gable, do we find a pronounced feeling for con-
struction. At the apex of a pyramidal group, generally somewhat
loosely compacted, rises the figure of Christ, which appears erect,
although in reality it follows the curve of the surface.
In the North, at a later time, church art transformed mass into
dynamic energy, which became a new source of quickened life. The
close alliance with the builder's art exerted a wholesome restraint
upon the rules of form ; in Italy this alliance was all too readily
dissolved, thus paving the way for an inartistic realism at times
pursued to a point at which it becomes positively repulsive.
Humanism then proclaimed itself dispenser of classicism and of
salvation.
One word more on the subject of line in its relation to mass.
The mischievous use of the term ' Gothic ', begun by Vasari,
shows how little we have yet learnt to appreciate the importance
of the Goths as carriers of culture. We are here concerned
primarily with the use of the line as a medium of expression.
To the representational South the human figure was a puppet,
its type selected, its attitudes, gestures, and features all arranged
for a play-acting mistaken for true expression. But in the North,
in Iran and Armenia, all effects are produced solely by means of
abstract form, by the use of lines at rest or in movement, in
harmony or in contrast ; the contribution of colour in enhancing
the effect of line cannot for the moment be defined. As examples
of linear expression we may quote the geometric devices of Islamic
art, derived from Iran, and the figures of animals in motion which
occur on early Northern bronze ornaments and on the panels of
the wooden church of Urnas (fig. 56, facing p. 153). In their use
2i6 INVESTIGATION OF ESSENTIAL CHARACTER
of line and colour the art of India and China are alike related to
that of the region between the Oxus and the Altai.
At this point it is particularly important to recall how strongly
the formal influence reacted upon the human figure as soon as it
was adopted by Northern art. Examples earlier than ' Gothic ' exist
in the south of France as well as in Armenia.^ The result is
seen during the great period of Northern Christian art when the
figure itself is neglected and the drapery becomes the real medium
of expression. A change first occurred when the North learned on
its own account to observe the human figure ; the movement was
taken up and developed in Italy, where it led to the study of the
nude and the reversion to classical models. Drapery in Egyptian
art lies stiffly upon the body ; in Greek art it sometimes clings to
the figure as if wet, after the Southern style, while at other times,
as in the North, it develops into a system of folds which sweeps
with ease and freedom across the body in the perfection of
rhythmic line.
B. Space. The artistic value of space, in its narrower sense
as a homogeneous interior, was first understood so far as Christian
art is concerned by the East-Aryans ; it is first found actually
K )Ci developed in Armenia. The original function of an interior space
as a place of assembly for the community was soon combined with
that of a house of worship. Thus spiritual unity corresponded
to the building standing free and compact, and might well have
found satisfaction in a dome over a central altar, a plan possibly
occurring in Armenia during the fourth century.
The building with lengthened nave first destroyed this unity ;
the genius of the architect who created the domed hall-church, or
that of a Vignola, was needed to restore it even in part. The hori-
zontal view focussed on the apse, as opposed to the upward view into
the dome, was at first the usual disposition in the Mediterranean
area, where it was associated with the triple division of the nave by
rows of columns. The provenance of the semi-circular apse and of
orientation is still an open question ; they were certainly first
definitely accepted in the interior of Asia Minor, and in Eastern
Syria and Armenia. The whole of the North gradually fell under
the influence of the triple-aisled type, and centralization was
sacrificed to length. Centralization, proceeding from Armenia,
only succeeded in making its way into Constantinople and the
^ The reader is referred to my book to compare the portals of Moissac, Autun,
on Armenia, ii. pp. 8ii f., and advised and other examples.
AND APPLICATION OF COMPARATIVE METHOD 217
architecture of the Orthodox Church ; here, by degrees, it entirely
superseded the long nave, though the struggle between dome and
apse continued.
Just as the facade destroyed the feeling for structural mass,
so the long nave was fatal to the feeling for interior space as
a compact unity. At first the basilica and the domed building
existed independently side by side, the former directing the eye
horizontally into distance, the latter upwards into height. Their
amalgamation gave rise to a conflict between centre and dome on
the one hand, and distance and altar on the other. In places
like Armenia, where the dome formed the starting-point in the
development of the communal church, its central position was
retained. But where the basilica was the original type, the
influence of the horizontal view forced the dome out of the centre
in the direction of the apse. It was the theologians of Nisibis
who succeeded in setthng the conflict, at least so far as representa-
tional decoration was concerned ; they retained the cross in the
apse as the symbol of sacrifice and redemption, and placed the
figure of Christ Pantokrator in the cupola, thus giving representa-
tional expression to the meaning of the dome.
The Christian was the first religious art to utilize interior
space for the expression of its constructional ideas. It is of course
possible that it was anticipated in this by some form of domed
temple in connexion with fire-worship or the cult of mysteries ;
nothing certain can yet be said on that point. But the function
of the church as a place of assembly was sufficient to ensure the
transference into religious architecture of an idea which had pre-
viously been confined to such secular buildings of general resort
as halls, baths, and palaces. The idea has but rarely been carried
to its simple and logical conclusion. The interiors of Armenian
and Aquitanian ' churches may be taken as models for the manner
in which the simple grandeur of their spatial eff"ect, unaided by
richer ornamentation, inspires in the visitor, as he crosses the
threshold, a sense of solemnity and awe, and awakens in him the
devotional mood. The sole way to this end is to treat the interior
space as a compact and homogeneous whole, and to heighten its
effect by a bold use of light and shade. This can only be done by
giving the architect control of the lighting of the interior ; but
the condition was not fulfilled after the fourth century. The
ecclesiastical movement very soon led to the closing up of the
* S. Front at Perigueux and its congeners.
2i8 INVESTIGATION OF ESSENTIAL CHARACTER
old large windows, and to a preference for mysterious gloom in
place of artistic effect. In the great period of Christian art in
Northern Europe the walls were once more pierced in order to
allow of the effective introduction of light and colour ; this
was, in truth, a return to the creative spirit of the first few
centuries.
A particular form of representation, the landscape, which
appears to have come from Iran together with the mosaic covering
of wall surfaces, was also instrumental, unconsciously perhaps, in
bringing into prominence the artistic value of space. The human
figures m the foreground, it is true, are grouped in a single plane,
and appear as if pressed against the front of the picture ; but in the
church of SS. Cosmas and Damian (Fig. 63 , facing p. 180) the clouds
descending from above, with the remoteness and elevation given to
the figure of Christ, together give the illusion of infinite space. The
effect of space suggested by the glitter of the gold background is
probably accidental, since its true value was purely decorative ; it
was merely another substitute for real gold, like those so commonly
used in India and Islam. In this particular mosaic the appearance
of depth is probably subjective and due to habit. The idea of
perspective m pictorial art was unknown to Iran ; its landscape
scenes are constructed by vertical projection without any attempt
at giving the illusion of depth. The figures of the Four Evange-
lists in S. Vitale are good examples of this kind of treatment. The
arrangement of zones in different colours used as a background in
later wall-paintings and miniatures was probably derived from
Hvarenah ideas.
Non-representational art was superseded by representation;
nevertheless its artistic values remained. To its influence was due
the change from the three-dimensional style of classical times
to the opposite mediaeval style. Early Christian art began this
important change by abandoning the illusion of space, and
developing representation along formal lines in two dimensions.
It almost seems as if the decorative impulse of the Northern and
nomadic races were forcing the plastic figures of Greek and
Semitic art forwards against the picture plane. The simultaneous
tendency towards a dramatic effect calculated to impress the
spectator finally stripped the human figure of all plastic sugges-
tion, and thus prepared the way for the true mediaeval style of
representation. The outstanding features of this style are the
strict frontality of the figures and their direct gaze towards the
AND APPLICATION OF COMPARATIVE METHOD 219
spectator.^ This does not mean that all feeling for nature and life
has been paralysed or overwhelmed by the gold background ;
it is simply the triumph of artistic form over imitation. The
artistically important point is the disposition of line and colour
upon a surface, not the use of subjects or figures for ecclesiastical
advertisement.
The best evidence for the withdrawal of the human
figure from nature lies in the rejection of even the limited
perspective admitted by classical art in favour of significance
in form. We see this in the disuse of true perspective, and the
arbitrary treatment of lines indicating recession. Hence arose the
so-called ' inverted perspective ', whereby figures in the middle
distance were commonly depicted on a larger scale than those in the
foreground. India and China did not originate this method, as
one might feel inclined to suppose ; they could only accentuate
it. Its true origin is to be sought in the Semitic point of view,
according to which the figure of the ruler or hero, later that of the
religious Founder, was regarded as transcending natural law.
Buddhism indeed needed no impulse from Graeco- Semitic art ;
in Sanchi and Barahat it was already following the same path.
Representational art in the West also remained more or less under
the influence of this style until the rise of scientific thought.
Exactitude in representation is not required by an art producing
its efi"ect from a distance ; indeed they are mutually incompatible,
a fact attested by numerous examples, including Raphael's ' Trans-
figuration '.
C. Light. In architecture lighting accords with the plan
of the building. In the timber-roofed basilica the windows in
the clerestory together unite with the aisle arcades to direct the
eye towards the apse ; in the domed building the drum with its
windows brings the lighting into harmony with the principle :
* Let the dome be central pomt and apex.' A peculiar method of
lighting is illustrated by the single-aisled barrel-vaulted building
and the hall church, which are lighted only from the sides.
In the history of development these tnree methods are not always
found alone ; sometimes we see a combination of two of them,
more rarely all three, as in Vignola's Gesu. Usually, lighting
from the sides is combined with one of the two types of lighting
from above. I believe that unity of effect in mass, space, and
^ This is a theme which I hope to lications of the Kunsthistorisches Institut
elaborate further in vol. xvii of the pub- connected with my chair.
220 INVESTIGATION OF ESSENTIAL CHARACTER
light was attained by the earliest Armenian domed structures of
the centralized type. Here the light from the drum windows
spreads upwards into the hemisphere above and downwards into
the apsidal niches, thus producing a tonal effect of perfect
rhythmic balance. In the basilica with upper windows only, the
intercolumniations produced a gentle gradation from the stronger
light of the central nave to the dimness of the aisles ; by develop-
ing this feature in conjunction with the barrel vault, very remark-
able solutions of lighting problems were achieved, especially
in the West and the North. Windows are sometimes found in the
west wall.
The use of light to relieve the monotony of mural surfaces
was originally confined to the East. In Armenia the domed
building was enlivened by means of a network of blind arcading,
afterwards reinforced by the shadow and light effects of vertical
triangular * slits '. At first the Mediterranean made no use of
such devices, but subsequently pilaster strips were introduced
from Mesopotamia, supplemented, at times, by arch mouldings
and later also by the true blind arcade. Finally Northern archi-
tecture (Gothic) wholly dissolved mass in tone values of free
space. In Italy an attempt was made to enrich lined walls by
tonal effects, using members of classical architecture ; but the
impression produced is little better than that of joinery. The
introduction of free-standing columns and genuine temple
fa9ades brought about a change ; but these organic members
harmonize neither with the storeyed building nor with the
dome.
Christian representational art, in so far as it models at all,
follows classical precedents ; it does not use its own powers of
observation, since in its representation it does not really perceive
nature. Shading grows weak ; it is barely suggested, and that
with timidity. It was left to Northern art to revive the free use
of tone-values ; the folds of drapery in ' Gothic ' statues are
hardly inferior to those of Greek sculpture, and sometimes even
more significant as pure form. Italian art carries on the Northern
tradition ; the gradual intrusion of the naturalistic figure is a
mark of decadence.
D. Colour. Surviving buildings prove that colour as a pre-
dominant value in interiors was born anew in the ' lined ' archi-
tecture of the East. The end was achieved by the covering
of walls with variegated slabs, and of floors and vaults with mosaic.
AND APPLICATION OF COMPARATIVE METHOD 221
How far the effect of light was heightened by fihration through
stained window-glass, we can only surmise in a tentative way
from the evidence of Islam and the West.^ In both of these pro-
vinces and especially in the North at a later period, the light
on entering the building was transformed into colour by passing
through richly coloured glass. The solid framework in which
it was set can still be observed in West-Gothic churches in Spain ;
it was derived from the similar stucco settings sporadically found
in Christian churches in the Balkans and, in wider distribution, in
the earlier mosques of the Mesopotamian type ; it would there-
fore appear to have originated in Persia. '^ The dim religious light
suffused with colour which is thus created in Northern cathedrals
does in cubic space what Rembrandt did on the picture plane.
There can be no question that these combinations of colour were
designed for definite effects. We know from historical records
that the interiors of Nestorian churches in China were so brilliantly
decorated ' that they resembled the plumage of a pheasant in
flight * ; some conception of what this means may be gained from
the choir of S. Vitale, in which the dazzling splendour of the
mosaics resembles the colours of a peacock's outspread tail.
5. Content. The basis of religion is a yearning of the spirit.
It leaves the individual soul unfettered, until the Church binds it
by her laws. The State exercises control in outward things ;
the Church differs from it only in its attempt to control the inner
life. It has perhaps been made clear that the spiritual content
of the formative arts at the appearance of Christianity differs
according to race and locality. The East- Aryan conceives a
centralized building over a square plan in which the vertical axis
dominates, the Semite a rectangular building in which the main
axis is transverse, the West-Aryan one in which the main axis is
longitudinal. It looks as though the latter merely adopted the
Greek temple plan with the image of the god at one end which
it replaced by a plain altar. The Hellenistic demand was
satisfied, in the structural sense, by the symbol of the apse which
made its appearance in Armenia, at first, it would seem, as nothing
more than a supporting buttress. True, it is certainly striking
that all Armenian churches without exception have their altars
towards the east end, and that early regulations are unanimous
^ In the excavation of S. Vitale, re- * The pictorial treatment of stained
mains of stained window-glass were found, glass finally eliminated the earlier
which may have been early Christian. method.
222 INVESTIGATION OF ESSENTIAL CHARACTER
in prescribing this direction. This may have been determined by
previous conditions, as I have already pointed out.
Christianity was twice introduced in the North, once before
and once after the great migrations ; on the first occasion it was
adopted by the Gauls from the Hellenism of the Mediterranean
area ; on the second, it reached the Franks from Rome. The
Goths were intermediaries between these two movements. They
brought the vaulted church from the Black Sea to the west of
Europe. Introduced by them in the Dark Ages, it eventually
triumphed over the types retained by Hellenism and Rome. The
spiritual significance of this intrusion of Eastern influence between
two successive periods of the timber-roofed basilica lay in its
disturbing and quickening power, destined inevitably to bear
fruit as soon as the religious feeling introduced into the North
took on a national tinge, and in the several countries there arose
personalities strong enough to pursue their own way and take
their inspiration wherever it was to be found.
In establishing his relations with the external world, the
early Christian sought to identify himself with the abnegation of
the Founder, who renounced every thought of temporal power
and possessions, an ideal only attainable far from the ordinary
affairs of life, not in it. As soon as he fell under the spell of
ordinary existence, succumbed to the desire for life, and con-
formed to the laws imposed on the community by Church and
State, Christianity as its Founder understood it ceased to exist.
To these changes correspond three successive motives influencing
the course of art : first, the prayer for intercession uttered by the
individual worshipper in the congregation ; second, the edification
of the community by the local churches ; finally, the exaction of
blind faith by the authority of the single state Church.
A. Intercession. The hope of redemption led to the pro-
duction of many works of art, sometimes complete churches,
sometimes parts of their decoration, which occasionally included
representations of the donors. The spirit prompting this work
was in harmony with that of Christ, but such work does not itself
appear to have grown up in a Christian environment ; it originated
in the religious thought of Iran, India, Greece, and Judaea, Purely
symbolical art, and art inspired by Hvarenah, or by Jewish funeral
and Paschal prayers, is confronted by representations of donors
with their names beneath, sometimes in large buildings executed
on such a scale as to be visible from a distance, in other cases set
AND APPLICATION OF COMPARATIVE METHOD 223
in places like Indian rock-cut chambers where they can scarcely
be seen at all ; we even find pavement mosaics completed in
sections contributed by many donors, and wall mosaics produced
under the same conditions, like those in the spandrels of the first
aisle in the church of S. Demetrius at Salonika. In this group
a popular substratum seems to reach the surface, but both Church
and Empire adopted and used it fully. The artist's freedom was
hampered by the usage of the particular country in which he
worked ; the conditions imposed both by patron and object
interfered with the expression of his individuality. Nevertheless
we see preserved by this Hvarenah tendency the spirit of those
Aryan artistic influences which contributed to the formation of
the early Christian style. Hellenistic and Mazdean art are alike
symbolical in expression ; the former takes for its symbol the
human figure ; the latter, animals, plants, and formal landscapes.
As vehicles of spiritual content they are essentially distinct from
the objective art of the Semites destined so soon to triumph over
both. The peculiar charm of the earliest Christian church art
during the predominance of the vault in the first four centuries,
as well as of such remnants as survived into later times, consists
in the fidelity with which it reflects the spirit of the Eastern and
Western Aryan races.
B. Instruction. In a movement initiated by the Church
for the definite purpose of instructing the faithful, and aiming
therefore at clearness in representation, little free play was per-
mitted to the artist. It is well known that types derived from pre-
Christian art were widely used as symbols universally understood.
They are given no new content ; the Church's sole demand was
that its meaning should be expressed without ambiguity. For
this purpose it issued rules and manuals for painters, which were
probably more common than present evidence suggests. A late
example has been preserved on Mt. Athos, but it has early proto-
types.
C. Use oj Dazzling Effects, National and popular art was sub-
merged under ecclesiastical authority ; ultimately a single power
prevailed in each half of the world, the Roman Church in the
West, and the Byzantine court in the East. Even the early
Christian state of Armenia could not permanently maintain its
individuality in the face of powers like these. The long-naved
building intruded upon the old domical type inherited from Iran ;
in like manner representation invaded the earlier formal art of
224 INVESTIGATION OF ESSENTIAL CHARACTER
ancient Mazdaism. In the sixth century all the provinces of the
empire were at the disposal of the Court artists, who were enabled
by the liberality of monarchs with a taste for architecture to
override local taste and tradition. They created nothing new,
but they did succeed in bringing out the latent possibilities of
what they borrowed. Natives of Rome or Byzantium appear
to have played only a subordinate part in this work ; witness the
church of S. Sophia and the artists who created it.
I am convinced that the nature of Eastern art was determined
at first by the individual character of national groups, whose
mentality was the joint product of environment and circumstances,
and early traditions, especially those of a religious nature. These
were the influences which prevailed in defiance of Church and State.
Amid all this clash of forces, as between dome, barrel-vault, and
wooden roof, square, broad and long building types, formal and
representational art, and their exploitation by Church and Court,
are we to take no account of the artist's individuality, which was
the outcome of his birth, education, position, and personal ideals ?
Does he not intervene with decisive effect at the chief turning-
points of development ? Can we cite no names of great personali-
ties who controlled the course of history in art ?
D. Artists. The particular occasions in the history of art
when names appear are in themselves significant ; thus we hear of
Zenobius, as the builder of Constantine's church of the Holy
Sepulchre ; of Anthemius of Tralles and Isidorus of Miletus as
the architects of Justinian ; and subsequently, in Armenia, of
the reactionary monk Manuel, and the progressive Trdat in the
service of the two Gagiks of Vaspurakan and Ani. All of these
men were court architects, commissioned by their rulers to
execute monuments of an exceptional kind. Are we to suppose
that religious impulse played any part in their creative work ?
Or were they solely inspired by the desire to surpass in size and
magnificence all previous efforts after monarchical advertisement ?
From the knowledge we have gained of East- Aryan art, and from
an independent critical study of what Eusebius has to say upon
the Church of the Holy Sepulchre, it is clear that none of these
architects created anything which was not ultimately derived from
the Iranian domed building on square plan. In all cases where
they were controlled by autocratic power, the early Christian
architects, like Leonardo and Bramante in later times, had recourse
to the dome, which in Iran had characterized the dwelling-
AND APPLICATION OF COMPARATIVE METHOD 225
house, and in Armenia the church. Thus a form of building
originally designed by a particular people for a particular purpose
was converted into a symbol of arrogant domination. The
artist no longer followed his own religious impulse, but obeyed the
dictates of authority. How far the Sassanians acted as pioneers in
this matter is a question into which I cannot now enter.
Or is it perhaps only the servile chronicler, seeking to find
favour with the ' hero ' of his time, who has preserved the names
of court favourites, and omitted all reference to the real innovators
in art, the men who had a national message to deliver ? Is pro-
gress in religious art conceivable without personality, or persona-
lity without a national background ? The general disappearance
of the vault which followed the rise of Christianity in the Roman
and Hellenistic culture areas, and the substitution of the timber-
roofed basilica, was perhaps due to the fact that officials
and not artists controlled art amid the medley of peoples
inhabiting the Mediterranean area in late Roman times. In the
East, where down to the fifth century the difl^erent communities
continued to go their own ways uninfluenced by Rome or Byzan-
tium, individuality could still make its power felt. I have
attempted, in a monograph on Armenia, to revive the memory of one
such splendid period of growth and energy in that country under
the Arsacid princes. Perhaps time will reveal an equally fascinat-
ing picture of creative rivalry in other parts of the East, such as
Mesopotamia, Syria, Asia Minor, and especially the central region
of Edessa and Nisibis. Wherever serious and intensive research
has been undertaken in Western art, as in the case of Lombardy,
France, and the Rhineland, a connected sequence of development
in church architecture has been demonstrated, to say nothing of
the logical development of Northern (Gothic) architecture from
the work achieved by mighty pioneers in later times.
The chronological sequence in Christian art is somewhat as
follows : first we find Iranian and Greek influences subsisting
side by side ; a Semitic influence then sets in, and ultimately,
radiating from Rome and Byzantium, sweeps all before it. The
conflict of these forces began before the Christian era. Greece
first took up the spiritual and intellectual contest with the Semitic
East, flowered, and succumbed. Imperial Rome only triumphed
by the aid of the Semitic elements which it derived through
Hellenism. The second attempt to overpower the victor in the
Mediterranean was begun by Christianity, aided by the spirit of
a45i g
226 INVESTIGATION OF ESSENTIAL CHARACTER
Northern Iran and the tradition of Praxitelean art. I believe
myself to have attained a clear view of the development in forma-
tive art. Jewish Christians with their non-representational style
found favour and support both among Mazdean Persians and
Greeks, receiving from each of these peoples an attractive sym-
bolical style. Greek and Persian might have joined forces.
Vaulted architecture and its decoration might have combined with
features of Hellenistic origin. But that would not have suited the
plans of the Church either in Rome or Byzantium, nor would it
have pleased the later courts which grew up in the North. The
first national flowers of Christian art, both Iranian and Greek,
were completely overwhelmed by Semitic influences from Edessa-
Nisibis and Jerusalem, the didactic tendency of which soon
surrendered in their turn to the interests of wealth and power.
The stronghold of these influences was Christian Byzantium.
The cheerful serenity of Hellenic art soon succumbed to the
spirit of domination ; Christ was no longer depicted as the
Shepherd and Spiritual Guide, but as the despot and the Judge.
The forms of Iranian architecture and decoration passed wholly
into the service of this anti-national movement. The path of
the student is beset with difliculties ; he is met, at the outset,
with almost pure Iranian art in Armenia, a similar Greek art
in Asia Minor, and Semitic art in Edessa-Nisibis ; turning
to the Roman mosaics and to Byzantium he perceives that
intermingling of influences which the men of the West and
the North, when they came South, marvelled at and copied
before they had had time to bring their own energies into play.
Northern love for interior space and decoration had already,
through the Goths, come to terms with the Iranian style ; but
the Semitically organized Church had at least as much power in
spiritual matters as Byzantium possessed in the East. The long-
naved church and the definite aim of representation were constant
obstacles to the free development of Aryan art, until city life gave
this art its chance, and from the surviving traditions of wooden
architecture it produced that supreme creation of northern genius
which is known as ' Gothic '. Italy eventually destroyed this
art, which had degenerated by a too histrionic treatment of the
human figure. She secured a triumph for the South, but this
misfortune did not occur until after the Northern blood in her
veins had produced a phase of representational art so brilliant that,
in some of its features, it bears comparison with that of ancient
AND APPLICATION OF COMPARATIVE METHOD 227
Greece. Nevertheless Leonardo, Michelangelo, and Giorgione
were defeated, like Diirer and Rembrandt after them. Personality
was almost entirely submerged in the conflict with a superficial
element inspired by Southern feeling, and regarding art as a pos-
session peculiarly its own.
I have already explained the plan of this attempt to investigate
essential character (p. 193). I have numbered the divisions as
follows : (i) Material and workmanship, (2) Subject, (3) Shape,
(4) Form, (5) Content. This sequence is rarely observed in
the actual history of development. It may be true that
Christian art in Europe followed this order between the
period of the great migrations and its climax in about a.d. 1500 ;
certain it is that the ' early Christian ' art which preceded it
reached its zenith almost immediately, and especially in the
fourth century. Its spiritual content (5) flowed from the perso-
nality of Christ ; this formed its starting-point and gave to all the
subjects, figures and forms which it adopted their essentially
Christian character. In the fifth and sixth centuries Church and
Court enter the field ; with them begins that exploitation (2) of
the germinal forms which they found in existence, and that
increase in size and magnificence which continued to the time
of Justinian, and often makes correct appreciation difficult.
Religious content (5) diminishes in proportion as the first four
elements increase in importance ; it only regains its full power
with the rise of a racial art in the North.
Thus principles through a whole millennium obscured by
the hot-house cultures of the South attained at last their full and
logical expansion on their native soil. To attribute this splendid
national movement to Southern influence or to scholasticism is
both unseemly and unpatriotic. An honest history of develop-
ment will be compelled to take account of the millennial persistence
of Northern principles, although it must be admitted that in the
end they were fully and brilliantly exploited by scholasticism.
In the foregoing chapters we have seen the North at work
in the South fashioning mediaeval art. In Greece and India
it yielded at first to the representational stream of the South ;
but in Parthia and Margiana, the ancient Sacian land, both
situated to the South of and between the Caspian Sea and the
Altai, it maintained its vital hold upon the people in spite of the
appeal which Semitic and Hellenistic art naturally made to the
material ambitions' of the Achaemenian and Sassanian rulers.
Q2
228 INVESTIGATION OF ESSENTIAL CHARACTER
The Arsacids carried Northern influence to Armenia. It is
important that the history of art should in future take account
of that triangular area, the base of which is formed by a line drawn
from the Caspian to the Altai, its apex by the mouth of the Indus.
Within this region the Hindu-Kush mountains deflected the main
stream of Indian culture towards Eastern Asia by way of the
Pamir and Khotan. This was the route followed by Buddhism,
Kabul forming the central point of the ' Gandhara movement ',
as did Ghasna that of Islam. The centres of Northern culture
were first Merv and Ferghana, later Bokhara and Samarkand,
while Herat lay on the frontier of a region full of originative power.
As a centre of intercommunication it is equally important for
the student of art and the historian of religion.^
More important are the problems of origins connected with
this region, to which I have made repeated allusion in this volume.
Not only the West, but also India and Eastern Asia, received, as we
have seen, streams of influence from this quarter ; but the art
which made the heaviest drafts upon the valleys of the Oxus and
Jaxartes is that which is generally named after its carrier, Islam.
The origins of Islamic art, even more than those of Western
mediaeval art, lie here. This region actually anticipated the
North as the source of a stream which in the course of a thousand
years overran the world. As regards the ecclesiastical provinces of
Margiana and Bactriana, with their bishop's sees at Merv and Balkh
respectively, and the more southerly centres of Herat and Sakastene,
the student should consult Sachau's book on the expansion of
Christianity in Asia ; he will then be as little surprised at the
appearance of the Nestorians in China as at the flood of silver
coinage from these regions in Northern Europe at the end of the
first millennium. The comparative study of art, with a view to
tracing the history of its development, has only been rendered
possible by the discovery of this Eurasian link connecting
^ For the religious field, the reader essential character upon the history of
may be referred to the attempt of development. It is to be hoped that his
W. Liidtke to explain the Recognition lead will soon be followed in the field
myth and the Placidas legend, and to the of formative art ; the importance of this
writings of K. von Spiess on the Fountain triangular region beyond the Oxus will
of everlasting youth. The researches of then be amply vindicated as a point of
Nathan Soderblom into the origins of junction for diflferent cultures. The
religion, published in 1915 under the works of Sir Aurel Stein have intro-
title of Das Werden des Gottesglaubens, duced us to a rich material from the
have shown how close is the bearing of Eastern side.
AND APPLICATION OF COMPARATIVE METHOD 229
Altai-Iran with the Indus. We may, perhaps, be permitted to
express the hope that the new world-order, which is even now
emerging, may find ways and means to send out expeditions to
excavate, and so acquire an accurate knowledge of this Asiatic
centre, which is of such extreme importance to the history of
civilization. For here Christianity and Buddhism met, the culture
of East Asia and our European North had their common home.
IX
Hiberno-Saxon Art in the Time of Bede
IN the preceding chapters I attempted to cover the whole field
of Early Christian and Mediaeval art ; in the present chapter,
written for the English edition of this book, I propose to re-
examine and supplement my work with special reference to the
most westerly section of this wide field. Does England tend
to confirm the view, which I have expressed with regard to
Christian art in general, that its early development can only be
properly understood if we enlarge our horizon, and learn to
appreciate not only the differences between popular art and the
art associated with Church and Court, but also national and local
differences in the East in Early Christian times and the relations,
in which formal and representational art stand to each other ?
I shall consider the problems in a certain sequence, which seems
to me to minimize the risk of overlooking points of importance.
Admittedly the framing of questions will in most cases be the
limit of my achievement ; my answers cannot be regarded as
an3rthing more than tentative suggestions.
I. The Study of the Monuments
In all the excellent English books on the subject of early
Christian art in Ireland, Scotland, and England, concluding with
Baldwin Brown's great work The Arts in Early England, the
student invariably finds the same uncertainty in the critical
appreciation of an art, which cannot be brought into connexion
with Rome or explained in the usual way by reference to Roman
monuments. No solution of the problem is possible so long as
modern investigators limit their horizon (as I myself did before
the publication of my Mschatta in 1902) to Syria and the great
Hellenistic cities. Baldwin Brown, in the second and fifth volumes
of his book, dealing with what I judge to be the most important
period, has collected all the threads of evidence with indefatigable
patience and care, and proved in a very convincing manner that
Rome did not exercise the decisive influence. This, however, is
HIBERNO-SAXONART 231
merely a negative conclusion, for which I may be able to substi-
tute something of a more positive character. By supplementing
my work of the last few years, summed up in this volume, with
the impressions formed on my journey through England in 1920,
I am able to venture upon a definite answer to the question of the
nature and origin of Christian art in England in the second half
of the first millennium. But, admittedly, the first thing to do is to
decide which monuments belong to the early period of Anglo-
Saxon Christianity.
Curiously enough, I find it necessary to express the hope that
the very existence of Anglo-Saxon monuments in England may
first find recognition. It is astonishing to see that Dehio and
Bezold's standard work on mediaeval art. Die Kirchliche Baukunst
des Abendlandes (1892), omits the extant Anglo-Saxon buildings
as of minor importance, and only uses Latin sources. Com-
prehensive study in this field of art is, in fact, a recent growth, and
one for which we are indebted to Baldwin Brown. His Ecclesias-
tical Architecture in England from the Conquest oj the Saxons to the
Norman Conquest (1903) is so thorough a piece of research, supple-
mented as it is by a catalogue and a chart of Saxon churches, that
such an omission as that of Dehio and Bezold is no longer con-
ceivable. Future students of the development of Christian art
will have to reckon with England as a distinct province, in the
same sense as the provinces discovered by me in Asia Minor,
Mesopotamia, and Armenia and, at an earlier date, by de Vogiie in
Syria. That, however, is only a beginning. In order to estabhsh
research on a firm foundation we require the determinant of time
as well as of place ; in this matter Anglo-Saxon buildings fare
no better than those of Asia Minor, of which scarcely a single
example can be dated with certainty. It was only by means of his
strictly scientific and comparative method, in contradistinction to
that of the philological historical school, that Baldwin Brown was
able to determine the existence of a distinct group of English
monuments in the seventh century at about the time of Bede.
In the province of sculpture we are scarcely more fortunate.
It was difficult to prove the existence of a flourishing period of
architecture in Anglo-Saxon times ; it is no less difficult to
obtain recognition of the fact that those important sculptures
preserved in the north of England, the Crosses of Bewcastle,
Ruthwell, and Hexham, date from the time of Bede. Such
recognition upsets the whole Roman scheme of chronology as
232 HIBERNO-SAXONART
it applies to England ; consequently repeated attempts have been
made to cast doubt upon the view which attributes these excep-
tionally important monuments to the earliest period of Northern
art. But it is one of the main principles of honest scientific
research that chronological sequences based upon local evidence
cannot be bodily transferred from one area to another. What
struggles has it not cost to free the dating of Christian monuments
in the East from the mechanical application of Roman standards ! ^
When I visited the Bewcastle Cross in person I learned with
some surprise, what the literature on the subject had never
informed me, that there is external evidence for attributing to it
a high antiquity ; this consists in the immediate proximity of an
extensive Roman fort, which makes it highly probable that there
was an early settlement in the neighbourhood. When we hear
from the philological side that there are no grounds for doubting
the authenticity of the historical portion of the inscription,
according to which the stone was cut in a.d. 670, we feel that we
can fairly accept what we are told without subjecting the date
to further special inquiry.
Nevertheless people might refuse to be convinced of the great
antiquity of the three High Crosses if the claims of Early England
were based on nothing more than these monuments and the
remains of her ancient buildings. But there is evidence of another
kind which proves beyond a doubt that the time of Bede was a
period of artistic achievement without parallel in the narrow zone
between Ireland and lona on the west and Northumberland and
Lindisfarne on the east; in other words, in the very area to which
the crosses belong. I refer to Irish illuminated MSS. such as the
Book of Kells and the Book of Durrow, and to Anglo-Saxon MSS.
under Irish influence, such as the Lindisfarne Gospels. Their
antiquity as Early Christian monuments is fortunately beyond
dispute, even for those who rely upon the unscientific philological-
1 Roman theologians may be ignored, sions will have to be made not only in the
since they are not to be converted ; but Christian East, but also at the extremity of
it is worth while to recall what happened the Christian West, in England. If it is
when Sir William Ramsay questioned my once clearly recognized that these crosses
early dating of the churches of Central date back to the seventh and eighth
AsiaMinorinthe.i4</ie«a«/OT, i903,p.656, centuries, it will be necessary to revise
and then, converted by his own excava- the existing history of art ; just as in the
tions, admitted the correctness of my East, such a recognition will open up
view in both an article in The Expositor quite new possibilities in the history of
(vol.iv, 1907) and in his book The Thou- development.
sand and One Churches. Similar admis-
INTHETIMEOFBEDE 233
historical method. Every handbook of art history informs us that,
to mention only the chief example, the Lindisfarne Gospels were
written a.d. 698-721 by Bishop Eadfrith. To many students this
is the first bedrock fact, from which scientific investigation into
the whole archaeological group must begin. In reality the study
of art must lay its own foundations firmly and independently
by continual use of the comparative method. This is a point
which I shall consider in greater detail.
II. Investigation oj Essential Character
Whether I turn to the remains of Anglo-Saxon churches,
to the crosses, or to the MSS. of the time of Bede, I feel in the
presence of an art which differs from every other art in the world.
Above all, I feel that we have here no mere pale reflection of Roman
or Byzantine art, but an art which, in comparison with these,
shows no less individuality than do the remains of buildings,
sculptures, and paintings of Hither Asia which I have been able to
identify as belonging to the Early Christian period. Anyone who
is familiar with the distant East — distant, that is, as seen from
Rome — feels himself here in the far West in a closely related
province. It will be our duty thoroughly to grasp the peculiar
nature of this Western art, just as we have grasped that of the
Christian East, and this can only be done by a systematic investiga-
tion of essential character on comparative lines. But it may not be
out of place to warn the reader at the outset against the common
fallacy of regarding imperial centres like Rome or Byzantium as
starting-points ; we found it more probable that such stages for
the scenic eflfects devised by Court, Church, and Academic Culture
do not create artistic values, but merely exploit them to advertise
their power. Remote and self-sufficing regions are much more
fitted to produce individuality in art than imperial centres, which,
like sponges, absorb everything within their reach.
I. Material and Work. It is beyond all doubt that in the
seventh and eighth centuries England contained a widely dis-
tributed series of architectural monuments. In these the
timber buildings customary in the North would appear to
have been superseded by stone structures which reveal traces
of their wooden origin no less clear than those observed in
Greek temples or in the railings of Indian stupas. The character
of the architecture that preceded the Persian period in Greece or
Asoka in India can be inferred from later buildings in stone ; and
234 HIBERNO-SAXONART
so it is in the case of England. Many of the towers so characteristic
of Early English architecture show us the wooden prototype trans-
lated into stone. The best example is Earls Barton (see Frontis-
piece) ; here the tower, when seen from a distance, has every
appearance of being timber-built, but a closer inspection reveals the
surprising fact that it really consists of rubble concrete faced with
stone ' beams ', in which the earlier wooden forms have been pre-
served with meticulous care. Even if the tower only goes back to
the later Anglo-Saxon period, it remains the most perfect and typical
example illustrating the origin of indigenous church architecture
in England. This class of building also shows the decorative
features characteristic of wooden construction in all periods : the
cushion capital, the knob capital with its surface rounded byturn-
ing,and the baluster. ThesefeaturesoccurinAnglo-SaxonEngland
no less than in Scandinavia, Georgia, Armenia, and Lombard
Northern Italy. In the building of enclosed rooms or halls the
whole of the North must be pictured as still in the wooden stage
at the commencement of the Christian era. We can hardly be
surprised at Bede's statement that building in the ' Scottish ' style
meant building in wood.
Nevertheless timber cannot have been the only building
material used by the Irish and Anglo-Saxons in Early Christian
times. In the crypts, which are the only extant portions of the
Anglo-Saxon cathedrals of Hexham and Ripon, the cutting and
the jointing of the stone are of such quality and so clearly the
result of long practice, that stone must undoubtedly be considered
as an alternative to wood in Early Christian architecture in
England. How otherwise can we account for the fact, not
appreciated by Dehio and Bezold, that noteworthy Anglo-Saxon
churches do not survive merely in Latin accounts with a Roman
colouring, but in their original stone and mortar ? Of wooden
buildings nothing remains, for Greenstead must be attributed to
a later period ; only chance or methodical excavation can still
bring to light remains like those discovered in Scandinavia at
Hemse or Oseberg. For the present we have to rely upon the
stone buildings, which are not mere imitations of their wooden
prototypes, but show clear traces of having developed peculiar
features independently of Roman architectural tradition.
Christian architecture might be defined as the construction
of interiors for the purposes of assembly and the ritual sacrifice.
So long as the architect clung to the traditional wood, so long as he
INTHETIMEOFBEDE 235
built only for his own generation with no thought of eternity, the
old indigenous methods prevailed. But as soon as he began to
build in stone, the vault made its appearance and with it the
characteristic feeling for organic construction. That the art of
vaulting was known in England in Bede's time is attested by the
subterranean chambers of Hexham and Ripon. They both con-
sist of a rectangular main chamber with barrel-vaulted roof, and
an antechamber attached to the west end ; in Ripon this is covered
with a transverse quarter- vault. At Hexham the roof adjacent to
the vault in front of the south staircase consists of stone slabs
inclined so as to form a gable. These are very important facts
and have never received the attention they deserve. The problem
which here confronts us is the same as that which reappears in so
perplexing a form in many of the Norman and ' Gothic ' cathedrals
of England : why was not the vault used to crown all that elaborate
substructure of walls and buttresses ? are we really justified in
ascribing such an eleventh-hour abandonment of plan to incom-
petence or timidity ? Is it not rather that Northern usage frustrated
the transition from a wooden roof to the vault, and explains the
delight obviously taken in heaping up huge masses of stone ? Even
in Palermo the Normans retained the wooden roof ; they could at
any moment have drawn on the resources of Byzantine architecture.
It might equally well be said that there was no one in Bede's
time competent to decorate manuscripts with representations of
biblical scenes in the manner of Rome and Byzantium, later
adopted in the European West under Charles the Great. Some
of the Evangelist figures in the Lindisfarne Gospels are unquestion-
ably derived from Greek prototypes ; other MSS. contain figures
of Evangelists and biblical scenes of similar origin. All these
figures are translated into the style of the calligrapher, who was
obliged to modify his old native methods to suit the new material
of parchment and the new craft of writing. The crosses show that
the stone-mason was much quicker to adapt himself to the
novelty of representing the human figure. It is an important
point that here, no less than in the MSS., the workmanship was
native, and not Roman as in the monuments of the preceding
period. Art was now founded upon nationality ; that is the out-
standing feature of the Christian movement. For the building of
basilicas it had been necessary to import foreign artificers from
Gaul and Italy. But Christianity in England, as in the East,
evoked forces not in the service of a foreign power such as Rome ;
236 HIBERNO-SAXONART
it tapped sources of national talent, hitherto exercised only in the
ornamentation of objects for daily use, and later of houses,
temples, or tombs. Here in the distant West the process was
closely analogous to that which I have demonstrated in the East.
2. Object or Purpose. All that survives of the architecture,
sculpture, and painting of the time of Bede was made for the
service of the Christian faith. Unhappily the works of art where-
with the Roman Church in England sought to create an impressive
display have perished in their entirety. But the remains of native
origin give the impression of an art freely practised and inde-
pendent of the dictates of ecclesiastical or any other kind of
authority. There is nothing servile in this national art ; each new
work proclaims Irish and Anglo-Saxon freedom from the in-
fluence of Rome. Culture in Bede's time had not been forced into
the grooves of tradition ; otherwise art could not have displayed
the freshness which distinguishes it. So far as architecture is
concerned, the reason for this independence must be attributed
to the failure of the Graeco-Roman triple-naved basilica to win
acceptance as a national form of expression. Great as was the
expenditure lavished upon these buildings by Roman missionaries
in the chief missionary centres, and imposing as they may have
been in outward appearance, they nevertheless occurred only as
sporadic examples. Here, in the North, the main purpose of
church buildings was from the very first a special one ; it was not
the few large city churches which gave the keynote, but the
numerous small buildings distributed all over the country, and
designed to protect the community in times of sudden attack.
Hence we find that even in wooden buildings all the customary
provisions for defence were made, as is proved by accounts of the
founding of the monastery at Lindisfarne. An indispensable
feature of this type of church throughout the North was a massive
tower, not built for height like the minaret or Italian campanile,
but resembling rather the ancient Indian monasteries in having
a number of inhabitable stories ; it often formed the central point
of the whole structure, which radiated from it like the arms of
a cross : excellent examples have survived both in England and
Sweden. Occasionally we also find the tower placed in front of the
long building. Another point of interest is the function of the
earliest crypts ; I have only seen those at Hexham and Ripon,
which have almost the appearance of habitable cells.
The customary interpretation of the crosses as gravestones
INTHETIMEOFBEDE 237
cannot be accepted without question. The character of the latter
is illustrated by the slabs of Hartlepool, those from Clonmacnois,
and others, which should be compared with examples from
Nestorian cemeteries in Central Asia. In addition to these,
examples occur, throughout the North, of memorial stones un-
connected with the grave cult ; in Scandinavia they still continued
to be erected and engraved with runes as late as the eleventh
century ; I observed similar, if isolated, examples in Armenia.
A monument of this type, supposed to be dated a.d. 922, and con-
sisting of two square shafts, is still standing erect at Odzun.^
Its Eastern face is covered with horizontal rows of figures, quite
contrary to Armenian custom, and its Northern face with orna-
mental designs ; it thus resembles the Ruthwell and Bewcastle
crosses. The biblical scenes, however, lack the modelling found
in certain ancient Georgian and Armenian churches ; they are
flat and treated in a manner quite different from the Southern style
so conspicuous in the figures of the English crosses. Here, too,
there is some doubt whether the monument is really a grave,* as
is generally assumed. The two obelisks have now been placed
next to the church underneath two arches approached by steps
(Fig. 65, facing p. 238). Possibly the two pilasters from Acre now
standing in front of S. Mark's at Venice (Fig. 68, facing p. 241)
belong to this group of stone monuments in pairs. I shall refer to
these again later.
The calligraphers were undoubtedly able to exercise freedom
of choice as between Roman and Eastern models. But it is not
unlikely that pressure was put upon them by Roman missionaries
to imitate the style of contemporary Graeco-Roman MSS., which,
whether in the form of the earlier papyrus rolls or of the later
vellum codex-pages, were illuminated with representational subjects
and never with abstract ornament in the true sense. Now one of the
distinguishing features of all Anglo-Saxon MSS. is the prevalence
of a homogeneous and purely artistic decoration. The placing
of the so-called Eusebian Canons at the beginning is significant.
The earliest dated example is to be seen in the Gospels of Rabula,
written in a.d. 586 in the monastery of S. John at Zagba in
Mesopotamia, and now preserved in the Laurentian Library at
Florence. The regular sequence of Evangelist's portrait and
initial is a matter of course in Armenian and Byzantine MSS.,
but not in Roman.
* Baukunst der Armenier, II. 695. * Of Ashot II.
238 HIBERNO-SAXONART
Particularly interesting is the so-called Franks casket in the
British Museum ; here a border of runes encloses a number of
carvings in low relief on whale's bone, the subjects being derived
in part from Northern mythology and in part from biblical history,
and that of Rome and Jerusalem. I expressed the opinion in 1904
that the latter or secular group probably goes back to the illustration
of Chronicles recording the world's history, like the Alexandrian
papyrus which I published in 1905. Bede does, in fact, allude to
an artistic cosmographical codex, which was given to King Aldfrith
by the Abbot Ceolfrith of Wearmouth and Yarrow (a. 0.690-7 16)
in payment for newly acquired land.^
3. Shape. In order to make himself readily intelligible,
the artist employs such shapes as are familiar both to him-
self and the spectator. Rome endeavoured to introduce the
Hellenistic basilica, with its triple nave and wooden roof, into
England as the regular church type. Did the indigenous Christi-
anity of the British Isles accept this alien type without question ?
Had it no other to offer in opposition ? In my book on Armenia
I have attempted to show how, in the Ararat district, a con-
structional unit originally borrowed from Iran became by
separate use and by enlargement the starting-point of a pecu-
liar development in Christian times. The England of Bede had
also its type distinct from the basilica. It has nothing to do with
the Armenian domed type on a square plan, but appears to be
descended from an earlier indigenous form, the single-naved long
building with or without a tower. It is certainly a noteworthy
fact that it existed in wide distribution in the seventh century.
Other special features of this building are the rectangular choir
and its separation from the public space by means of walls or
supports projecting at right angles from the sides. The same
features have been noted in the wooden churches of Sweden, and
have been explained as survivals of ancient Northern temples.
If that is really the case, and Christianity in the North Sea area
originally adopted the shape of an earlier pagan temple, it should
be possible to prove the existence of buildings directly descended
from it, and independent of the Roman basilica. In Armenia also
the type occurs as a religious building at the same period as
in England ; subsequently in Provence and Aquitaine it became
the starting-point of an important development in form. The
^ The allusion is in the Anonymous Lives of the Abbots ; see Plummer, vol. i.
p. 380, par. 15.
' \ ! / ^
n
65 Odzun ; pillars from monument See p. 237.
ON
N
O
c
c
a,
c
e
V
1-1
v
x
i
I
«
u
u
<
INTHETIMEOFBEDE 239
question is whether England was not the pioneer country for this
type of building in the West, just as Armenia and Mesopotamia
were in the East. The central tower may have led to the develop-
ment of the cruciform plan as early as the time of Bede, although
the existing examples are all later. In Eastern Christianity all
these single-naved types were the source from which the art of
vaulting sprang. Did not the same thing occur in England ?
Subterranean vaults existed in Bede's time ; in Monkwearmouth
we also find an example of barrel vaulting on the ground-floor
of the tower. No definite conclusion is possible on this point at
present ; nor can we be certain whether its origin was indigenous
or connected with Rome or the East.
The problem of the origin of types in Anglo-Saxon art recurs
in an accentuated form in the case of the human figures and
decorative motives. If, as Baldwin Brown has proved, these were
not derived from Rome, what was their origin ? Are they indi-
genous, or is a connexion with other provinces of art conceivable ?
We should naturally expect that the decorative motives were
indigenous. It is beyond question that much of this rich fund of
figures is Celtic-Irish and much also Teutonic and Anglo-Saxon
in origin ; it is unnecessary to refer once more to particular details.
But there remains a considerable residue of motives as to which the
problem arises whether they were borrowed as chance suggested
from one or other of these sources, or whether they are all derived
from some third common source. My investigation of this pro-
blem will be based upon the high crosses and the manuscripts ;
/or the sake of brevity I shall ignore the decoration of buildings.
In the Acca cross of a.d. 740 (Fig. 66, opposite) three of
the surfaces are ornamented with vine-scrolls, rising in a series
of interlaced circles or pointed ovals, in which the natural form of
the vine is wholly subordinated to the geometric design. The
chief school for this type of ornament is, as I have shown, the
Mshatta fa9ade ; there it is possible to trace its evolution in all its
stages, varying from the realistic to the absolutely conventional.
The ' rows of pellets ', which appear at the upper end of the
Acca cross, were also a favourite form of decoration in Altai-
Iran and the East. In the Bewcastle cross (Fig. 67, facing
p. 240) three of the surfaces have scroll designs, the construction
of which is reminiscent of the Mshatta fa9ade in that the
stems spring directly from the corner or the centre of the base
line without the intervention of a vase. The manner of their
240 HIBERNO-SAXONART
interlacing has features which reappear in a well-known monument
from the region of Antiochene influence ; I refer to the pilasters
from Acre now in Venice (Fig. 68, facing p. 241). Comparison does
indeed reveal that the work of the Antiochene artist was further
from the Eastern and nearer to the Graeco-Roman style than was
that of the Anglo-Saxon . The reason for this is simple ; the Anglo-
Saxon and the Iranian, in virtue of their Northern affinities and
remoteness from Rome, are in closer agreement with each other than
with the Aramaean of Palestine, who, in his advanced position on the
Mediterranean coast, was led to vacillate between Iranian-Northern
conventionalism and the naturalism of the Greeks and Semites.
In my Altai-Iran (p. 204 f.) under the heading ' multiple
surface ' (Mehrfldchigkeit) I brought together for comparison
Afghan and Norwegian examples of wood and stucco ornament in
angular relief ; I then considered pierced work ; both methods
were treated as characteristic of nomadic and northern art. In
introducing the * Nigg Stone ', a slab with a cross in Ross-shire,
I would first call attention to stelae, such as those of Armenia,
sometimes of enormous dimensions, erected as memorials or grave-
stones ; above all I would remark upon the angular relief given
to the lines of the sculptured designs on both sides and the pierced
bosses on one side of the plinth (Fig. 69, facing p. 242) ; here the
cross stands out in the front plane above a sunk background, an
effect reproduced in colour in the Lindisfarne Gospels. The
ground is filled with scrolls, the negative residue of which, reserved
in champleve style in the stone, is fashioned into round or
lozenge-shaped bosses covered themselves with interlacings pre-
senting sharp contrasts of light and shadow. The reader should
compare with this the stucco frieze at Deir es-Suryani by the
Natron Lakes in Lower Egypt, executed, as I believe, by Iranian
workers in stucco for the Abbot Moses of Nisibis in the tenth
century (Fig. 70, facing p. 243). Apart from the general resem-
blance in the feeling of the ornament, I would call special attention
to the horizontal frieze below and the quartered bosses enclosed
in the lower scrolls.
In the purely decorative pages of the Irish-Anglo-Saxon
manuscripts one peculiarity is particularly striking. I refer to the
ornamental projections — the humanist would explain them as
Akroteria — ^which appear in the four angles and still oftener in the
middle of the sides (Fig. 71, facing p. 244) ; these have attracted
even less attention, so far as I am aware, than the existence of
CC
U
05
o
C3
3
1^
o
68 Pillars from Acre in Venice. See pp. 237, 240.
INTHETIMEOFBEDE 241
the ornamental pages themselves. Can slab-crosses like the
Nigg Stone have furnished the model for these ? Or indeed for
the whole idea of a framed surface completely filled with orna-
ment ? The reader may compare with these the bordered
panels of the Kairwan Mimbar, which I believe to be derived
from Iran (Fig. 35, facing p. 117). Such works of art, inspired
by the inmost feehngs of the nomadic and northern races and
executed in a definite material with a definite technique, fur-
nished the prototypes from which, in some as yet unknown
region of Hither Asia, was born the new style of ornament-
ing parchment ; for the purpose of reconstructing this style
the manuscripts of early England are no less important than
Armenian and Byzantine illuminated books of later date. It
originated in another region than the Merovingian fish-and-bird
initials, which never travelled as far as England. This is a signi-
ficant fact, since it is clear proof that the Anglo-Saxon province
of art was independent of the Prankish. Where then was the
home of the decorative art enriching Anglo-Saxon manuscripts in
the time of Bede ? At home, or in the East ?
I believe that the clue to the problem is to be found in those
peculiar salient motives, introduced as an ornamental feature in
the angles and on the sides of those * frames ', which are a general
and independent product of the Northern spirit, quite distinct
from the architectural panel borders of the Southern culture-
area. These projecting motives appear in the East in conjunction
with the rn-shaped moulding described in my books on Altai-
Iran and Armenia. The Armenian miniature (Fig. 40, facing
p. 123) shows that we have here to do with a form of art
which resembles the Middle Byzantine in its insistence on this
feature. The reader is referred to my remarks on this subject in
Kleinarmenische Miniaturenmalerei.
It is very striking to find on the Bewcastle and Ruthwell
crosses not only scroll ornament, but also representations of
human figures ; these occur on one side of the former, and on
two sides of the latter cross. Though they have no connexion
with Rome, they strongly suggest the sarcophagi of Ravenna,
and show the closest agreement, as I pointed out in my book on
Armenia (ii. 720), with those ancient flat reliefs, which we some-
times find no less surprising on Armenian and Georgian churches
than in Anglo-Saxon England. Possibly we have here a branch
spreading westwards, as another had spread eastwards, from the
3451 K
24a HIBERNO-SAXON ART
same Antiochene stem formed by the convergence of Hellenistic
Asia Minor and Semitic Mesopotamia. We are only just begin-
ning to observe these relationships ; hitherto all eyes have been
blinded by the Roman obsession. In spite of the necessity for
being brief, I should like to recommend a comparison between the
figure of Christ in the cruciform church at Mzchet (Fig. 72, facing
p. 245) in Georgia with the two figures of Christ on the Bewcastle
and Ruthwell crosses, so frequently illustrated together. It is as
though the same hand had been at work, not only in the posture,
bearing, and features, but also in the schematic treatment, a subject
to which I shall refer again later. The figure would be the more
striking if, as Baldwin Brown remarks, it had been depicted
without the full beard and with the moustache only. The reader
is advised to read my comment on this subject in Altai-Iran
(p. 262 f.). He should also compare these figures with the Gan-
dhara type of Buddha.
It has often been observed that the introduction of the
falconer in the Bewcastle cross points to a direct influence from the
East. Perhaps we should also include the figure of the archer,
which appears on the terminal cross of the Ruthwell stone and
corresponds to the four figures of the Evangelists which must
have existed on the back. The archer occupying the lower limb
of the cross, we may perhaps supply on the others the unicorn,
the mountain, and a bird, somewhat as we find them in a landscape
of Iranian style in the MS. roll of Ku-K*ai-Chi in the British
Museum, where the archer kneels over against the landscape.
What is here represented is the struggle between good and evil,
a Mazdean motive which started in Iran and during the Middle
Ages is found everywhere between China in the East and England
in the West.
4. Form. In demonstrating that a variety of influences and
prototypes may have determined the course of an artistic move-
ment, 1 am preparing the way at the end of which their independent
action will become intelligible. What, now, is the peculiar factor
determining form in Anglo-Saxon art ? The bulk of the surviving
monuments reveals an unpretentious treatment of interior space
in architecture, and in sculpture and painting a preference for pure
decoration. Representation in formative art and the use of the
vault in building are both of secondary importance.
On buildings there is apparently no trace of the excessive
ornament employed in Norman work. The churches are almost
69 ' Nigg-stone.' See p. 24c.
o
CO
c
<3
C/2
1>
Q
i>
s
o
<i
iS
c
i>
S
e«
C
O
o
3
+-*
C/3
O
INTHETIMEOFBEDE 243
entirely undecorated, as they were in Syria and Armenia, where in
hke manner the earliest builders were intent on construction pure
and simple. The national stone architecture of the earlier
Christian centuries in Armenia eschewed elaborate ornament.
The exterior produces its effect by the lines of its compact and
massive structure ; the interior space in the one-aisled indigenous
churches is unique in its proportions. In what has preceded, I
have not emphasized the problem of origin ; in the present place
I must make it clear that a type perhaps derived from the pagan
temple may have developed into a fertile form in the hands of
Christian builders as early as the time of Bede. It is significant
that at so early a period England should have possessed a type
peculiar to herself : in the other parts of Europe, Dehio and
Bezold have established the existence of the one-aisled type only
after a.d. iooo. In England, as in Armenia, the type stands at the
beginning of the whole development, and, if a Georgian reviewer
of my work on Armenia is correct,^ it was also the primitive plan
in Georgia, though here there was a groined vault over the middle.
In England this single-aisled church has proportions which at
once arrest the attention (Fig. 73 , facing p. 246). The interior space
looks as if it had been heightened to accord with the adjoining
tower. At any rate the height is so remarkable in comparison
with the breadth that we are reminded of the proportions in
Gothic cathedrals, where the breadth is to the height as i to 3I.
A tendency to vaulting is intrinsic to the single-aisled building ;
I cannot believe that it was resisted by Anglo-Saxon architects.
The impulse towards vaulting is plainly expressed in the design
of the later Norman churches, though even in the case of three-
aisled buildings this vaulting was not finally carried out. But
it is remarkable that here there should have been a feeling for
organic development which was wholly lacking in the case of the
basilica. The point of departure for this development is always
the vault, for which due supports have to be provided. In the
North other influences led up to it, derived from wooden construc-
tion ; this style of building was perhaps more fitted than that
of the Graeco-Roman basilica to combine the feeling for organic
construction with the three-aisled plan. In this connexion I throw
out the following conjectures.
I assume the existence in early England of wooden churches
in addition to those built of stone. Arguing from a parallel case in
* Tschubinaschwili, 1921.
R 2
244 HIBERNO-SAXONART
Norway, I infer that side by side with the simple stone churches
there arose a wooden type, in plan and elevation closely related to
the basilica, I mean the so-called ' mast churches '. In these the
nave is raised on tall masts above the level of the aisle-roofs, thus
enabling windows to be placed in the upper walls and admitting
direct light to the whole middle part of the church. It has
generally been supposed that this treatment of walls is later than
the vaulted Romanesque churches, and that it is to be explained as
an imitation in wood of a basilican type in stone. As a matter of
fact, as I have already pointed out (Figs. 14-17, facing p. 80),
the wooden churches of the North nave little to do with
Romanesque, so little that they probably introduced the develop-
ment of Gothic themselves. Among other things I may point to
their decisive influence in originating the triforium, a member which
people have usually associated with the Southern and Byzantine
gallery. Viewed in this connexion, the triforium would represent
the first effort to break the monotony of the nave walls, raised to an
excessive height through a custom which had become traditional,
as we have seen, in single-naved churches built of stone. A glance
at Fig. 18 shows the sequence : at a given height the rows of tall
masts required strutting. Even in wooden construction the
requirement was in part fulfilled by means of arcading. In stone
churches, and as an enrichment of upper walls, this arcading
remained as a decorative feature when its practical function
had ceased.
Anglo-Saxon art of about a.d. 700, from whatever point we
view it, is not antique but mediaeval in its forms, so purely
* mediaeval ' that in pre-Carolingian times we can only find its
like in two quarters : in the North in so far as it was uncon-
taminated by southern influence ; and in the Asiatic East,
though here only in the so-called period of antiquity. It is a con-
sequence of this that while in architecture pure space was the aim,
in the other arts all idea of depth is lacking, and likewise all idea
of the effect to be produced by light and shade. The only features
that tell are surface line and colour. If figures enter into the plan
they look exactly as if thrust forward from behind against the
picture plane ; ornamental motives have no definite direction ;
everywhere there is a preference for the repeat-pattern. These
distinguishing marks are at least as characteristic of Anglo-Saxon
as of any other art of the North or East. All centres in ornament.
If in the time of Bede reliefs appear (or are we to say reappear }),
yi Page of the Lindisfarne Gospels, fol. 26'^ ; British Museum. See p. 240.
72 Mzchet ; the Christ outside the apse. See pp. 242, 245.
INTHETIMEOFBEDE 245
they submit to the constraint of the design in every case where
they are free of southern influence.
The reHefs upon stone crosses show certain peculiarities also
shared by monuments in a Greek style contrasting with that of
late-Roman sarcophagi, with their undercutting and its eff^ects
of light and shadow ; the monuments to which I refer are found
on Islands of the Aegean, and in isolated instances in Georgia and
Armenia. I take as an example the figure of the founder in the
church of Mzchet (Fig. 72, opposite). Compare the Christ of
the Bewcastle Cross. The ground is cut away from the border, as
in champleve work, and Christ in his rigidly frontal pose reserved
in such a manner that the ridges of the complicated folds in the
drapery are kept uniformly at the level of the border.
The above-mentioned sequence of the full-page illuminations
in books of the Gospels is worthy of especial remark. First comes
a page of pure ornament, then the portrait of the Evangelist,
finally the beginning of the text in large initials. Single pages
with crosses occur in all oriental manuscripts, whether Syriac,
Coptic, Armenian, or Byzantine : under Islam, these decorative
pages, though of course without the cross, are especially frequent
in the oldest MSS. of the Koran. But the usual sequence at the
beginning of the book, whether Irish or Anglo-Saxon, appears to
represent a native development actuated by Eastern suggestions.
In the play of line in these ornamental pages and in the initials
we note a quite exceptional delight in personal invention. Tame
indeed is the ornament of Byzantine MSS. in comparison, where
the decorative page is generally contracted to a mere head-piece,
as seen in an illustration from a Paris MS. of Gregory of Nazianzus
(Fig. 74, facing p. 247). All this joy in decorative design must have
come from the scriptoria of Hither Asia.
Turning the pages of Anglo-Saxon MSS. I am again and again
struck by the recurrence of the colour-combination of red and
yellow. For me this in itself is enough to establish the close con-
nexion of the group with Coptic illumination, though the Coptic
work is infinitely cruder. Wherever you have to deal with the
art of nomadic or northern peoples you find these two colours
dominant ; there are certain classes of objects, notably carpets and
embroideries, in which the colour scheme never goes beyond this
point. At present I am unable to follow this clue further for lack of
published monographs on the subject bringing together sufficient
material to justify general conclusions. It may be that in illumina-
246 HIBERNO-SAXONART
tion the colours are connected with the material, parchment, to
which they are applied.
5. Content. The spiritual content of any art is the first of its
values to be considered, even though its development is known.
We may say at once of Anglo-Saxon works of art that they create
an impression of spontaneity and freshness as enduring as that
produced by works of early Greek or early Gothic art. In the
values employed at the time and in the country of Bede, and in
their combined effect, there is the attraction of youthfulness and
frank simplicity, qualities which may be regarded as more original
than those of the Roman art which preceded them or the Norman
art which came after. Since the impression created by different
works is not the same, the crosses and the MSS. should be con-
sidered as separate groups ; the determining factor lies not merely
in the culture represented, but even more in the national in-
dividuality evinced. This is the feature in which the art of earlier
and later times was more or less deficient. The individuality of
the Anglo-Saxons, like that of the Franks and Lombards, was
somewhat repressed through contact with southern culture and
through the effect of their conversion to Christianity upon their
purely northern style. At the time of Bede we clearly mark the
transition from a decorative to a didactic and representational art ;
the luxuriant design of the MSS. belongs to the days before
theology ; the northern joy in linear dexterity persists by the side
of the foreign intrusive elements.
We should observe the extraordinary difference in tempera-
ment between the artists who executed the decoration in the Book
of Kells and the Lindisfarne Gospel. We might take the one for
a Celt, the other for a German.
III. History oj Development
We can put the matter in a few words by saying that Anglo-
Saxon Christian art developed no less independently of Rome than
did that of the various countries of Hither Asia. Native elements
were naturally preferred, but it need excite no surprise that
oriental motives were welcomed in England by virtue of their
analogy to those of northern art.
Professor Baldwin Brown's work, The Arts in Early England,
so admirable in its methodical treatment of Anglo-Saxon buildings,
so thorough in its investigation of their character and evolution,
reveals the necessity for testing the current myth as to the course
73 Bradford-on-Avon Church ; interior. See p. 243.
if? ■■;
>%-j .-.^jCTOCtnOH-TTTlXp^
1 ituju K-or.w
ri
— ro U tirjj c i ? " totouj
j^ di O 4to P * aucr ' '
iwiimmMxmzuW:
74 Heading of MS. Paris Bibl. Nat. 543, fol. z%\ See p. 245.
INTHETIMEOFBEDE 247
followed by development in the field of early Christian architec-
ture as a whole. This myth came into existence because points of
small evolutional significance, or even of no significance at all,
were put into the foreground ; in the demarcation of the antique
and the mediaeval periods, and, again, in relation to Vasari's
division between Gothic and Renaissance art, points were treated
as cardinal which were in fact of secondary importance.
Baldwin Brown rightly perceives that the basilica ' is
strangely lacking in principles of growth ', and that the mediaeval
movement could no more have started from basilican building
than from Italy and Rome. He finds in Gaul, and later in the
North, from the sixth to the ninth century, new social forces at
work to re-design the long-naved church in combination with the
centralized type. In England this attempt may be traced in the
employment of both these constructional forms at Hexham.
Professor Brown finds in the description of the Hexham churches
analogies with the Octagon built by Constantine at Antioch, as
it is described by Eusebius.^
The essential point is to be found in the negative admission that
the timber-roofed basilica contained no principle of growth ;
it lacked the power of organic development. I may add a sup-
plementary argument of a more positive kind. This power of
development existed in Hither Asia ; it was bound to make itself
felt as soon as the step from wooden building to stone construction
was taken in the North : this is a point which Professor Baldwin
Brown has overlooked. The first effects ensued in the country
churches. The large cathedrals of Anglo-Saxon episcopal cities
and the great monastic churches are all unfortunately lost ; but
though they freely used columns in the Hellenistic manner, it
may be assumed that they also made tentative efforts towards true
organic construction, because there were regions in which pioneer
construction of this kind had already appeared.
Similar observations may be made in the case of high crosses.
These do not connect with the Roman tradition interrupted about
A.D. 450; if they did, the connexion would be unmistakable,
however rough the work. As, however, despite the interval of
two hundred years, these crosses rank with the best work produced
during a period of half a millennium, we may perhaps infer not
only the creative vitality of the local element, but also the
importation into the island of some vigorous foreign influence
^ The Arts in Early England, ii, pp. 320 fF.
248 HIBERNO-SAXONART
such as that which I beheve to have been exerted in Gaul
and the Prankish dominions by the Visigoths and their retinue
of Anatolian and Armenian craftsmen. The parallel examples
from the Orient which I was able to adduce do in fact suggest
that the general trend of influence in those times from the East
westwards was decisively felt in England. We have here merely
to inquire how such a penetration became possible.
For the sake of brevity, I have hitherto said nothing of the
earlier art which flourished on English soil. We discern its
character in the linear designs on the bronze ornaments of the La
Tene period, slant-cut and thus heightening the effect by a play
of light and shadow over the bright metal ; ^ we discern it too
in intrusive Teutonic features with their analogy to Scandinavian
art, and the motives which the Lombards and Franks brought with
them into Italy and Gaul. For the entry of oriental influences into
England two routes were available, the one by land, the other by
sea. More than once we have found traces suggesting a connexion
with Antioch. I refer especially to monastic types reaching
Ireland from the East Mediterranean by the usual maritime routes
followed by trade and colonization. These types were developing
in Ireland during the decisive years when the Anglo-Saxons were
conquering England and were preserved by a national Church
throughout the two centuries when Teutonic paganism estab-
lished itself in that country. Does any other hypothesis explain
the fact that the Roman mission of Augustine and his successors
was promptly confronted with a formidable resistance, formidable
because it was offered not by paganism, but by an advanced
culture of East-Christian origin ? Does any other hypothesis
explain the appearance on the Continent of Irish monks as teachers
of Greek in the reign of Charles the Great } How, if we reject it,
are we to account for the Gospels of Lindisfarne produced on that
English lona half a century after the foundation of its monastery
in A. D. 632 ? For the luxuriant illumination of this book reveals
no single trace of Roman influence, but incontrovertible proofs
of all those germinative forces, native and oriental, the character
of which I have attempted to demonstrate in these pages. The
discovery in Northumberland of a Greek inscription relating to
a Syrian from Commagene suggests to me a confirmation, external
and material if you will, of influence from the theological centre
^ For the origin of such methods see in Zeitschrift fur bildende Kunst, 1921-
my Altai-Iran, and compare my paper
INTHETIMEOFBEDE 249
of Nisibis transmitted through Antioch. The oriental origin of
Archbishop Theodore, a native of Tarsus, points in the same
direction; so, finally, does the manner in which Cassiodorus
adopted the theological system of Nisibis in his monastery of
Vivarium, setting an example for all later monastic foundations
and for the whole mediaeval culture of the West.
The second route, that by land, ran from the Black Sea
across Russia to Scandinavia. It is characteristic that recent
research ignores its very existence.^ The reluctance of scholars
used to the old ways to give it serious consideration is intelligible
enough ; an understanding of its significance throws too clear
a light on the consecrated myth that all development comes
from Rome ; the debility of the whole tradition is too sharply
exposed.
When Mazdaism had forced back Semitism and taken con-
trol of Hellenism, the North entered upon a period of independent
development. What we call Gothic is the third art of northern
origin which rose to greatness by its own merits. The Greek
genius grew great in the South, the Iranian in the borderland
between North and South, the Germanic in the North itself.
In these pages we are concerned with the transition between
the first and second of these achievements, during which, in so
far as Christian art is concerned, I have shown Armenia to have
been a main creative centre ; the almost contemporary expansion of
Islam on Iranian soil I here leave out of account. Throughout the
centuries before the rise of Gothic art it was the function of Western
Europe to act as a kind of mirror for all movements coming from
the East ; as early as the La Tene period, and more obviously yet
after the Anglo-Saxon invasion, we find the characteristic signs of
the exercise of this function. The North Sea area, previously far
behind that of the Baltic, which was of decisive importance for the
earlier Indo-Germanic migrations, now took the lead : interlaced
bands, beast-ornament, the curious diagonal repeat-pattern treated
by English stonemasons and illuminators in the same way as
polygonal designs at a later time in Mohammedan art, finally
initials in MSS. — these all travelled to and fro by land from one
end of the North to the other. At present the source of the initials
which dominate the ornament of manuscripts is not certainly
known. I incline towards one connected with Iran, possibly the
Manichaeans.
* Compare my Altai-Iran, and Arne's writings.
250 HIBERNO-SAXONART
We cannot as yet definitely trace the course of the strong
Antiochene and Armenian influences apparent in Anglo-Saxon art ;
enough that they exist. Their recognition will quicken the activi-
ties of research and heighten the satisfaction in the study of our
own Northern art. One piece of evidence strikes me as deserving
of attention. In the Libellus Islandorum there occur in a list of
Icelandic bishops the names of three Armenians, Peter, Abraham,
and Stephen.^ The MS. in which this list occurs dates from the
middle of the twelfth century ; it was probably first written in
Latin and later translated into Anglo-Saxon ; the Armenian
bishops probably belong to the eleventh century, and must
certainly be placed before 1125. How did they reach Iceland?
By the northern or by the southern route ? And are these
orientals only found in the North Sea area, or had they predeces-
sors, perhaps as far back as the time of Bede ?
IV. The Subjective Side
With regard to one aspect of artistic inquiry I have left the
student to read between the lines ; I refer to the aspect affecting the
spectator himself, as opposed to the objective side of the monuments,
their artistic character, and the evolutional forces which find utter-
ance through them. I have said nothing of the attitude of aesthetic
criticism towards Early Christian art, though from the outset I
opposed the usual standpoint of scholars who assign the leading
place to Rome and to a Christian classical art. Perhaps I may be
permitted, in conclusion, to say a few words as to the manner
in which this error has affected research in the field of Anglo-
Saxon art.
In the second volume of his Lombardic Art Rivoira gave
a classical exposition of the Roman theory. If, he asks, Britain and
Ireland had not been devoid of native artists, would Benedict
Biscop and Wilfrid have introduced foreign craftsmen ? The
North, needless to say, is regarded as incapable of developing
unaided such an art as that which flourished in the time of Bede.
Facts count for nothing against sophisticated Latin sources, or
against that tissue of imagination spun by Church, Court, and
Humanism together during the last few centuries, and undis-
turbed in its traditional academic supremacy by any illumination
from geographical, historical, or economic sources. In the case of
^ Communicated by Charles Singer. Origines Islandicae, i, Oxford, 1905,
Cf. G. Vigfusson and F. York Powell, p. 299, note 2.
INTHETIMEOFBEDE 251
Rivoira, who was my highly respected friend, we have also to
reckon with the influence of patriotic feeling for Italy, and above
all for Rome, which inevitably warped his judgements.
We are now beginning to write the history of art-history ; the
one chance of improvement lies in a knowledge of the path already
traversed. As long as we believe in the myth still repeated in the
handbooks, there can be no prospect of freeing research from the
dross of a fixed tradition. Who can say whether he has really
stepped clear of antiquated beliefs and stands now upon firm
ground ? The present book has only been concerned with a short
period of two centuries in the first Christian millennium. Our
study can only attain its definitive form by rising to a survey of
the whole field in which its aims and activities are displayed.
But no one nowadays has the leisure, whatever the ripeness of his
supposed knowledge, to accomplish a task so immense as this.
When, a good many years ago, I defined my position with
regard to the proverb ' all roads lead to Rome ', people chose to
see in my action a political movement in favour of emancipation
from Rome. To-day I am suspected of raising the ' battle-cry of
the new aesthetic ' the moment I confront representational
Southern art with a purely decorative art of the North. With the
problems of beauty my line of inquiry has for the time being
nothing to do. The present volume is absolutely unconcerned
with them ; it is occupied with questions of essential character and
of development without pretending in any way to pass judgements
from the emotional or individual standpoint of the spectator. It
is natural enough that there should be people who prefer the Acca
Cross to the crosses at Ruth well and Bewcastle, because its
sculpture is composed of pure ornament without any admixture
of edification in the form of Christian figure-subjects. For us the
one important thing is to point out that in the time of Bede there
actually existed monumental crosses which on no less than three
of their sides renounced representation in favour of vine-scrolls
geometrically planned.
There can be no doubt that artistically the British Isles belong
not to Rome but to the North, and, their eastern part more
especially, to the North Sea area. In the second half of the first
millennium their relations with the North were closer than with
the Continent immediately opposite their shores, whence the
Roman mission under Augustine came. Ireland, with Scotland
and Northumbria, felt both the northern influences coming from
252 HIBERNO-SAXONART
Eurasia by the land route and the southern influences arriving by
sea. The former resulted from the movement discussed in my
Altai-Iran und Volkerzvanderung,^ the latter from the development
which formed the subject of Die Baukunst der Armenier und Europa.
Both together contributed to the remarkable Northumbrian art of
Bede's time, which seems to be even more important than the
pagan art of the North in Scandinavia, and beyond, as far as the
seed-bed of Islamic art between Altai and Iran. We have to
envisage the whole of this vast region as a single unit. We shall
then be able to understand the immensity of the blunder made by
a contemporary northern scholar whose eyes are for ever fixed
on Rome, whether it be the imperial city of the first half-millen-
nium after Christ, or the papal city of the second. Through this
error he denies his own country, and her part in a northern
artistic province in its degree at least as worthy of attention as the
southern province in Hellenistic times, or the Armenian province
on the borderland between South and North. Science has nothing
to do with this or that artistic preference, this or that conception ;
it demands of us one thing and one only, that we renounce our
predilections, that we become the faithful instruments of facts,
and that we leave the facts to speak for themselves. Only if we so
act shall we cease to move round and round the old circles of the
theorists ; only so shall we find the solid support of a development
discovered by comparative research, and of a history based on a
specialized and scientific knowledge. The first pre-requisites of
this reform are a horizon which, theoretically at least, embraces all
the world, and a point of view which does not suffer the obtrusion
of particulars to obscure the vision of the whole.
^ The reviewer of the German edition Iran. Otherwise he could hardly have
of the present book in the ' Literary suggested that I had paid too little atten-
Supplement' of The Times (June i6th, tion to the work of RostovtzefF.
1 921) seems to have overlooked Altai-
BIBLIOGRAPHY
THE following list of works, arranged in chronological order,
forms a continuation to my Orient oder Rom, 1901, Kleinasien, ein
Neuland, 1903, Koptische Kunst, 1904, and the last section in
vol. iii of Byzantinische Denkmdler, 1903, p. 119 f. (cf. periodical reports
in the Byzantinische Zeitschrift). This list only includes such of my
publications as bear upon the subject-matter of this volume.
1903. Seidenstoffe aus Agypten im Kaiser-Friedrich-Museum. Wechsel-
wirkung zwischen China, Persien und Syrien in spdtantiker Zeit.
(Jahrbuch der preussischen Kunstsammlungen, xxiv, 1903,
pp. 147-78, with 19 illustrations in the text.)
1903. Antinoe-Bawit und die deutsche Wissenschaft. (Supplement, All-
gemeine Zeitung, 12. September 1903, no. 206, pp. 493-5.)
1903. Der Ursprung der ' romanischen ' Kunst. (Zeitschrift fur bildende
Kunst, N.F. xiv, pp. 295-8 ; 3 illustrations.)
1903. Der angebliche Stillstand der Architekturentvxicklung von Kon-
stantin his auf Karl den Grossen. (Zeitschrift fiir Bauwesen,
liii, 1903, pp. 629-34.)
1903. Der Pinienzapfen als Wasserspeier. (Mitteilungen des deutschen
archaologischen Instituts in Rom, xviii, 1903, pp. 185-206,
with 13 illustrations.)
1904. Der Dom zu Aachen und seine Entstellung. Ein Protest. Leipzig.
Hinrichs. 1904.
1904. Mschatta, Kunstwissenschaftliche Untersuchung. (Jahrbuch der
preussischen Kunstsanmilungen, xxv, 1904, pp. 225-373 ;
4 plates and 104 illustrations in the text.)
1905. Die Schicksale des Hellenismus in der bildenden Kunst. (Neue
Jahrbiicher fiir das klassische Altertum, Geschichte und deutsche
Literatur, xv, 1905, pp. 19-33, with i plate and 4 illustrations
in the text.)
1 905 . Die christliche Kunst in einigen Museen des Balkans . (Osterreichische
Rundschau, iii, 1905, Part 30, pp. 158-65.)
1905. Bine alexandrinische Weltchronik ; Text und Miniaturen eines
griechischen Papyrus der Sammlung W. Goleniicev. Adolf Bauer
und Josef Strzygowski . (Denkschrift der Akademie der Wissen-
schaften in Wien, philosophisch-historische Klasse, li, 1905.)
1906. Die Miniaturen des serbischen Psalters der koniglichen Hof- und
Staatsbibliothek in MUnchen. Nach einer Belgrader Kopie
ergdnzt und im Zusammenhange mit der syrischen Bilderredaktion
des Psalters untersucht. With an introduction by V. Jagid.
254 BIBLIOGRAPHY
(Denkschrift der Akademie der Wissenschaften in Wien, philo-
sophisch-historische Klasse, Hi, 1906.)
1906. Spalato, ein Markstein der romanischen Kunst bet ihrem Ubergange
vom Oriente nach dem Abendlande, (Studien aus Kunst und
Geschichte, Friedrich Schneider gewidmet von seinen Freunden
und Verehrern. Freiburg i. Br., Herder, 1906, pp. 325-36 ;
6 plates and 10 illustrations.)
1906. Review of Walter Altmann's Die romischen Grabaltdre der Kaiser -
zeit. (Gottingische gel. Anzeigen, 1906, no. 11, pp. 908-14.)
1907. Kleinarmenische Miniaturenmalerei. Die Miniaturen des Tubinger
Evangeliars MA. XIII, i vomjahre H13 bezw. 893 nach Chris ti
Geburt. (Veroffentlichungen der koniglichen Universitatsbiblio-
thek zu Tubingen, ii, 1907 ; 4 plates and 11 illustrations in the
text.)
1907. A Sarcophagus of the Sidatnara type. (Journ. Hell. Stud, xxvii,
1907, p. 99 f.)
1907. Chrtstliche Antike. (Supplement to the Mvinchener Allgemeine
Zeitung, 1907, no. 64, pp. 505-6.)
1907. Bildende Kunst und Orientalistik. (Memnon, i, 1907, pp. 9-18.)
1907. Zum Christustypus. (Der Tiirmer, ix, 1907, part 10, pp. 505-9.)
1907. Der Kiosk von Konia. (Zeitschrift fiir Geschichte der Architektur,
i, 1907, pp. 3-9.)
1907. Amra als Bauwerk. (Zeitschrift fiir Geschichte der Architektur,
i, 1907, pp. 57-64, with 3 illustrations.)
1907. Amra und seine Malereien. (Zeitschrift fiir bildende Kunst, N.F.
xxviii, 1907, pp. 213 ff., with 6 illustrations.)
1908. Das orientalische Italien. (Monatshefte fiir Kunstwissenschaft,
i, 1908, pp. 6-34, with 12 illustrations.)
1908. Zur friihgermanischen Baukunst. (Zeitschrift fiir Geschichte der
Architektur, i, part 10, 1908, pp. 247-50 ; i illustration.)
1908. Muhammadan Art. (Hastings' Encyclopaedia of Religion and
Ethics, i, pp. 874-81.)
1908. Altchristliche Kunst. (Die Religion in Geschichte und Gegenwart,
i, pp. 381-97.)
1908. Oriental Carpets. (Burlington Magazine, Ixvii, 1907, pp. 25-8.)
1908. Neuentdeckte Mosaiken von Salonik. (Monatshefte fiir Kunst-
wissenschaft, i, 1908, pp. 1019-22 ; 2 illustrations.)
1909. Die Geburtsstunde des christlichen Kirchenbaues . (Supplement,
Miinchener Allgemeine Zeitung, 1909, no. 51, pp. 417-19.)
1909. Der signalformige Tisch und der dlteste Typus des Refektoriums.
(' Worter und Sachen ', i, 1909, pp. 70-80 ; 10 illustrations.)
1909. Antike, Islam und Okzident. (Neue Jahrbiicher fiir das klassische
Altertum, 1909-10, xxiii, part 5, pp. 354-72 ; 19 illustrations.)
1909. Die persische Trompenkuppel. (ZeitschriJft fiir Geschichte der
Architektur, iii, 1909, pp. 1-15 ; 13 illustrations.)
BIBLIOGRAPHY " 255
1910. Wilperts Kritik meiner alexandrinischen Weltchronik. (Romische
Quartalschrift, xxiv, 1910, pp. 172-5.)
1910. Der Eintritt Mesopotamiens in die Geschichte der christlichen Kunst.
(Monatshefte fiir Kunstwissenschaft, iii, 1910, pp. 1-4 ; with
5 illustrations.)
19 10. Die nachklassische Kunst auf dem Balkan. (Jahrbuch des freien
deutschen Hochstiftes zu Frankfurt am Main, 1910, pp. 30-43.)
1910. Amida. Mate'riaux pour Vepigraphie et Vhistoire musulmanes du
Diyar-Bekr par Max van Berchem. Beitrage zur Kunstge-
schichte des Mittelalters von Nord-Mesopotamien, Hellas und
dem Abendlande. With a supplement, ' The Churches and
Monasteries of the Tur Abdin ' by Gertrude L. Bell. (23 plates
and 330 illustrations. Heidelberg. 1910. Karl Winter.)
191 1. Orientalische Kunst in Dalmatien. (Bruckner, * Dalmatien und das
Osterreichische Kiistenland '. Lectures delivered in March
1910, on the occasion of the first expedition of Vienna Uni-
versity. Vienna and Leipzig. Deuticke. 191 1.)
191 1. Felsendom und Aksamoschee. (Der Islam, ii, 191 1, pp. 79-97;
with 5 plates.)
igii. Kunsthistorisches in H. Grothe, ' Meine Vorderasienexpedition
1906-7 ', 1910, pp. ccxii-ccxxviii ; with 15 illustrations in the
text and one map.
igi I. Kara- Amid. (Orientalisches Archiv, i, 191 1, pp. 5-7; with
2 plates and i illustration in the text.)
191 1. Ornamente altarabischer Grabsteine in Kairo. (Der Islam, ii,
pp. 305-36 ; with 38 illustrations in the text.)
191 1. Das Problem der persischen Kunst. (Orientalistische Literatur-
zeitung, xiv, 191 1, col. 505-12.)
1911. Ein zweites Etschmiadzin-Evangeliar . (Huschardzan, Festschrift
aus Anlass des loo-jahrigen Bestandes der Mechitaristenkon-
gregation in Wien, pp. 345-52, 191 1 ; with 5 illustrations in
the text and 2 plates.)
1912. Der grosse hellenistische Kunstkreis im Innern Asiens. (Zeitschrift
fiir Assyriologie, xxvii, 1912, pp. 139-46, with i plate.)
1913. Alhambra. (Enzyklopadie des Islams, i, 1913, pp. 292-4.)
1913. Ein Werk der Volkskunst im Lichte der Kunstforschung. (Werke der
Volkskunst, i, 191 3, p. 12 f.; with i plate and 5 illustrations
in the text.)
1913. Ostasien im Rahmen vergleichender Kunstforschung. (Ostasiatische
Zeitschrift, ii, 1913, pp. 1-15-)
1914. Erworbene Rechte der osterreichischen Kunstforschung im nahen
Orient. (Osterreichische Monatsschrift fiir den Orient, xl,
1914, pp. 1-14 ; with 9 illustrations.)
1914. Zentralasien als Forschungsgebiet. (Osterreichische Monatsschrift
fiir den Orient, xl, 1914, pp. 68-82 ; with 18 illustrations.)
256 BIBLIOGRAPHY
191 5. Die Entstehung der Kreuzkuppelkirche . (Zeitschrift fiir Geschichte
der Architektur, vi, 1913, pp. 51-77 ; with 26 illustrations.)
1915. Das Mausoleum Konig Karots von Rumdnien. (Osterreichische
Monatsschrift fiir den Orient, xli, 1915, pp. 46-8 ; i illus-
tration.)
1915. Die sasanidische Kirche und ihre Ausstattung. (Monatshefte fiir
Kunstwissenschaft, viii, 1915, pp. 349-65 ; with 5 illustrations
in the text and 17 figures on 8 plates.)
1916. Der Ur sprung des trikonchen Kirchenbaues. (Zeitschrift fiir
christliche Kunst, xxviii, 1916, pp. 181-90 ; i plate and 10
illustrations.)
1916. Ravenna ah Vorort aramdischer Kunst. (Oriens christianus, N.F.
V, 1915, pp. 83-110 ; with I plate and 8 illustrations.)
1916. Ornamentet hos de altaiska och iranska folken. (Konsthistoriska
Sallskapets publikation, Stockholm, 1916, pp. 5-18 ; with
II illustrations.)
19 1 6. Die hildende Kunst des Ostens. Ein Uberblick iiber die fiir Europa
bedeutungsvollen HauptstromuTigen. 28 illustrations. Leipzig.
Werner Klinkhardt. 1916. (Bibliothek des Ostens, ed. Kosch,
vol. iii.)
1916. Religion och personlighet i den bildande konsten. (Ord och Bild,
XXV, 1916, pp. 625 ff. ; with 12 illustrations.)
19x7. Altai-Iran und Volkertoanderung. Ziergeschichtliche Untersuchun-
gen iiber den Eintritt der Wander- und Nordvolker in die Treib-
hduser geistigen Lebens. Ankniipfend an einen Schatzfund in
Albanien. 229 illustrations and lo plates. Leipzig. Hinrichs.
1917-
1 9 17. Discussion of P. Frankl : Die Entwicklungsphasen der neueren
Baukunst, and A. Schmarsow : Kompositionsgesetze in der Kunst
des Mittelalters. (Theologische Literaturzeitung, 41, 1916,
col. 61-4, and 513-16.)
1918. Die bildende Kunst der Arier. (Deutsche Warschauer Zeitung,
nos. 59 and 60-1, 28 Feb. and i and 3 March 1918.)
1918. Vergleichende Kunstforschung auf geographischer Grundlage. (Mit-
teilungen der Geographischen Gesellschaft in Wien, Ixi, 1918,
pp. 20-48 and 153-8.)
191 8. Die Baukunst der Armenier und Europa. Ergebnisse einer vom
Kunsthistorischen Institute der Universitdt Wien 1913 durchge-
fUhrten Forschungsreise, planmdssig bearbeitet. 2 vols. Vienna.
A. Schroll & Co. 1918.
19 1 8. Persischer Hellenismus in christlicher Zierkunst. (Repertorium fiir
Kunstwissenschaft, xli, 1918, pp. 125-48 ; with 10 illustrations.)
1918. Leonardo-Bramante-Vignola im Rahmen vergleichender Kunstforsch-
ung. (Mitteilungen des Kunsthistorischen Instituts, Florence,
iv, 1918, pp. 1-37 ; with 12 illustrations.)
B IBLIOGRAPHY 257
19 19. Probleme der nordischen Kunst. A lecture delivered to the * Kunst-
historische Gesellschaft ' at Stockholm. (Konsthistoriska Sall-
skapets publikation, 1919.)
19 19. Suden und Mittelalter. (Monatshefte fiir Kunstwissenschaft, xii,
1919, pp. 313-23.)
1919. Norden und Renaissance. (Zeitschrift fiir bildende Kunst, xxxi,
1920, Jan.)
1920. Neue Bahnen der Kunstforschung. (Osterreichische Rundschau,
Ixii, 1920, pp. 33-8.)
1920. Ein Christusrelief und altchristliche Kapitelle in Bulgarien. (Byzan-
tinisch-neugriechische Jahrbiicher, ed. by Nikos A. Bees I.)
1920. Orpheus- und verwandte iranische Bilder, (Contribution to Otto
Kern : Orpheus, eine religionsgeschichtliche Untersuchung.)
WORKS OF COLLABORATORS
Ernst Diez :
1906. Die Funde von Krungl und Hohenberg. (Jahrbuch der k. k. Zentral-
kommission fiir Denkmalpflege, iv, 1906, pp. 202-27 ; 2 plates.)
1910. Bemalte Elfenbeinkdsten und Pyxiden der islamischen Kunst. (Jahr-
buch der preussischen Kunstsammlungen, xxxi, 1910, pp. 231-
44 ; I plate and 4 illustrations.)
1914. Isfahan. (Zeitschrift fiir bildende Kunst, xxvi, 1914-15, pp. 90-
104 and 113-28 ; 34 illustrations.)
191 5. Die Kunst der islamischen Volker. (Akademische Verlagsgesell-
schaft Athenaum. Berlin-Neubabelsberg. 1915. 5 plates and
288 illustrations.)
1916. Burgen in Vorderasien. (Der Burgwart, 1916, pp. 90-101 ; 10
illustrations.)
1918. Churasanische Baudenkmdler . Vol. i. (Berlin. Dietrich Reimer,
19 1 8. 5 colour and 36 other plates, and 40 illustrations.)
Vol. vii of the works of the ' Kunsthistorisches Institut ' of
Vienna University (Lehrkanzel Strzygowski).
L. POTPESCHNIGG :
1912. Aus der Kindheit bildender Kunst. Leipzig. B. G. Teubner.
(Saemannschriften fiir Erziehung und Unterricht. Part 2.)
191 5. Einfiihrung in die Betrachtung von Werken der bildenden Kunst.
Wien. Schulbiicherverlag. 1915. Works of the ' Institut '.
Vol. ii.
19 1 7. Planmdssige Wesensforschung in der Dichtkunst. (Neue Jahrbiicher
fiir das klass. Altertum. Leipzig, 1917, vol. xl, part 2,
pp. 209-34.)
J45» S
258 BIBLIOGRAPHY
Richard Kurt Donin :
1915. Romanische Portale in Niederosterreich. (Jahrbuch des Kunst-
historischen Instituts der Zentralkommission fur Denkmal-
pflege.) Works of the * Institut '. Vol. iv. Vienna. Anton
Schroll & Co. 1915.
Heinrich Gluck :
1 9 16. Ein islatnisches Heiligtum auf dent Olherg. Beitrag zur Geschichte
des islamischen Raumbaues. (Der Islam, iv, part 4, p. 328.)
19 16. Der Brett- und Langhausbau in Syrien, auf kulturgeographischer
Grundlage bearbeitet. With 49 illustrations and 4 plates. (Zeit-
schrift fiir Geschichte der Architektur, Beiheft 14. Heidelberg.
C. Winter. 1916.) Works of the * Institut ', vol. vi.
1917. Die beiden ' Sasanidischen ' Drachenreliefs, Grundlagen zur seld-
schukischen Skulptur. With 5 plates. (Publications of the
Musee Imperial Ottoman, part 4. Constantinople. Ahmed
Ihsan & Co., 1917.)
19 17. Tiirkische Kunst. Lecture delivered at the session of the Hungarian
Institute of Science at Constantinople, 5 May 1917. (Mitteilun-
gen des ungarischen wissenschaftlichen Institutes in Konstan-
tinopel. 1917. Part i. With 26 illustrations. Budapest-
Constantinople.)
1920. Das Hebdomon und seine Resie in Makrikoi. Contributions to the
comparative study of art, pubHshed by the ' Kunsthistorisches
Institut ' of Vienna University (Lehrkanzel Strzygowski).
Part I.
1920. Die Bdder Konstantinopels und ihre Stellung in der Baugeschichte
des Morgen- und Abendlandes. (Vienna, 1921.) Works of the
' Institut ', vol. xii.
Artur Wachsberger :
1916. Stilkritische Studien zur Wandmalerei Chinesisch-Turkestans . (Ost-
asiatische Zeitschrift. Special publications, no. 2. Berlin.
Osterheld & Co., 1916.) Works of the ' Institut ', vol. iii.
Karl With :
19 19. Buddhistische Plastik in Japan bis zutn Beginne des S.Jahrhunderts.
Ergebnisse einer 19 13-14 vom Kunsthistorischen Institute der
Wiener Universitdt [Lehrkanzel Strzygowski) nach Ostasien unter-
nommenen Forschungsreise. Kunstverlag Anton Schroll & Co.,
19 19. Works of the * Institut ', vol. xi.
1
BIBLIOGRAPHY 259
The following works of the Kunsthistorisches Institut of the University
of Vienna (Lehrkanzel Strzygowski) have been pubHshed since 1920 :
Vol. iii. K. Ginhardt : Das christliche Kapitell zwtschen Anttke und
Spdtgotik. Vienna, 1923.
Vol. iv. J. Strzygowski : Kunde, Wesen, Entwicklung. Mit Beitragen
von E. IDiez, K. Ginhardt, H. Glvick, T. Plutzar, M. Stiassny,
E. Wellesz. Vienna, 1922.
Vol. xiv. M. Dimand : Die Verzierung der koptischen Wollwirkereien.
Stromungen des Weltverkehres im Kreise der Mittelmeerkunst.
Leipzig, Hinrichs, 192 1.
Vol. xvii. H. Berstl : Das Rautnproblem in der altchristlichen Malerei.
(Liithgen's ' Forschungen zur Formgeschichte der Kunst '.) 1920.
Vol. xviii. S. Kramrisch : Wesen der altindischen Kunst. (This will
probably appear in Liithgen's ' Forschungen zur Formgeschichte
der Kunst '.)
Vol. XX. J. Strzygowski : Die Krisis der Geisteswissenschaften. Vienna,
1923.
Vol. xxii. C. Petranu : Ifthaltsproblem und Kunstgeschichte. Vienna, 1921.
Vol. xxiii. E. Diez : Persien, Islamische Baukunst in Churasan. 1923.
Vol. XXV. Josef Strzygowski : Die indischen Miniaturen im Schlosse
Schonhrunn. Vienna, 1923.
Studien zur Kunst des Ostens, Josef Strzygowski zum sechzig-
sten Geburtstage von seinen Freunden und Schiilern. Edited
by Heinrich Gliick. Vienna and Hellerau, 1923.
s 2
INDEX
Acca cross, 239.
Achthamar, 63, 123, 145, 150, 159, 160,
167, 171.
Acre pilasters, 113, 237, 240.
Adam and Eve, 160.
Adiabene, 53.
Agnellus, 85, 135.
Ainaloff, D., 4, 38.
Aix-la-Chapelle, Cathedral of, 90, 200,
202.
Ajanta, 139.
Akathistic Hymn, 160.
Albanians, 142.
Alberti, 98.
Albigenses, 90, 126.
Aleppo, 60, 89.
Alexander, 2, 14, 173, 175.
Alexandria, 33, 37, 38, 40, 54, 109,
115, 163.
Altai, III, 229.
Ambrose, St., 84.
Amesha Spenta, 174.
Amida (Diarbekr), 6, 23, 32, 70, 125,
135. 140-
'Amman, 112, 113, 124.
Ampullae, 167.
'Amra, see Quseir 'Amra.
Angora, 147.
Am, 65, 82, 99, 145.
— Cathedral, 72, 90, 94.
Antioch, 25, 33, 37, 38, 40, 85, 140,
141, 163, 165, 172, 177, 247, 248.
Apse decoration, 136 fF., 171, 213.
Aquinas, Thomas, 206.
Aquitaine, 87 ; see also Perigord, P^ri-
gueux, S. Front.
Ara Pacis, 151.
Arab influence, 9, 10, 26, 39, 67.
' Arabesque ', The, 112.
Arabs, 31, 131, 188.
Aramaean influence, 14, 31.
— cycle, 166.
Aramaeans, 31, 161, 162.
Arbela, Chronicles of, 24, 25.
Arcading, Blind, 89, 109-10, 112, 116,
121.
Arch, 145.
— Horseshoe, 87, 148.
— Pointed, 89.
Architectural types, 54, 61 fF,
Armenia, 3, 7, 8, 15, 17, 26, 36, 55,
58, 6i if., 68, 99, 127, 142, 144, 175,
188.
Armenian influence, 46, 71, 72, 81,
III, 112.
Armenians, 32, 33, 82, 85, 133.
Armenian churches, 36, 37, 160, 179,
200.
Armenian MSS., 127, 135 ; see also
Achthamar, Ani, Bagaran, Thalin,
Thalish, Zwarthnotz.
Artik, Cathedral of, 141.
Artists, 224.
Aryan influence, 20, 22, 8i.
Aryans, 11, 32.
Asia Minor, 33, 39, 40, 56.
Asoka, 18, 79, 173, 2J3.
Assembly Hall, 200 n.
Asterius, 40.
Athos, Mount, 161, 169, 170.
Augustine, 248.
Auvergne, 88.
Avesta, see Zend.
Bactria, 23.
Bagaran, Cathedral, 63.
Baghdad, 147.
Baeratids, 131.
Baldwin Brown, 230, 231, 246, 247.
Balkh, 228.
Baptism, 163.
Baptistery, The, 58, 134.
Barahat Stupa, 157, 219..
Baroque, 203.
Barrel-vault, 22, 37, 55, 67 fF., 81, 86 fF.
— decoration, i34fF.
Basil I, 170.
Basilica, 3, 36, 37, 38, 54, 86, 183, 197.
— Domed, 70, 71.
— Ursiana, 135.
' Basket ' capital, 147.
Baths, Roman, 77.
262
INDEX
Bazaklik, 123.
Beast- Symbolism, 121.
Bede, 231 ff.
Bethlehem, 45, 123, 163, 183.
Bewcastle Cross, 231, 232, 239, 241 ff.
Bezold, Dehio, 231.
Bisutun, 147.
Bodhi-tree, 158.
Bokhara, 228.
Borgund, Church at, 96.
Bramante, 3, 98, 99, 100.
Brick building, 8, 105.
— Sun-dried, 60, 160, 194.
British Museum, 165, 238, 242.
Brunelleschi, 98.
Buddhist art, 35.
Bukovina, 160.
Bulgaria, 89, 114; see also Philippo-
polis, Patleina, Pirdop, Sofia.
Bull capitals, 147.
Burckhardt, 170.
Bus-i-Hor, iii.
Byzantium, see Constantinople.
Cairo, 107.
' Candelabrum,' 123 ; see also Tree,
Formal.
Capital, see cushion, ' knob,' &c.
Cappadocia, 40, 68.
Capital, development of the, 145, 146,
210.
Carinthia, 201.
Casaranello, 134.
Cashmere, see Kashmir.
Cassiodorus, 53, 162, 205, 249.
Catacombs, Art of the, 15, 115, 130,
159, 166, 169.
Caverns, rock, 166.
Chair of Maximianus, 151, 165.
Chalcedon, Council of, 26, 132, 142.
Chambord, 99.
Charlemagne, 84, 184.
Chase, Scenes from the, 119, 122, 150.
Chess, 174.
China, 18, 23, 24, 78, iii.
Chinese landscape, 119, 120.
Chojak, 147.
Chrysotriclinion, 133.
Church, Influence of the, 28, 29, 35 ff.,
41,42,46, 48, 198.
Church types ; see Basilica, Cruciform
domed church.
Cilicia, 99, 146.
Cimabue, 170.
Cividale, 114.
Classical art, Christian, 13, 16, 20, 33.
Clonmacnois, 237.
Cologne, 92.
Colour, 220 ff.
Column, The, 77, 145.
Columns, triumphal, 157.
Commacini, 85.
Constantine, 18, 33, 42, 45, 56, 168,
17?, 200, 247.
— Churches of, 45, 46, 56.
— at Antioch, Octagon of, 38, 45, 66.
— Porphyrogennetos, 133.
Constantmople, 15, 26, 33, 40, 44, 46,
48, 127, 134, 146, 170, 172, 182, 195,
208 , 209 . See also Holy Apostles , Nea,
S. Irene, S. Sophia.
Content, 221 ff.
Copts, 31.
Costume, Mazdean, 125.
Councils, 131 ; and see under Chalce-
don, &c.
Court, Influence of, 15, 30, 42, 43, 46,
48, 53, 100.
Cracow, 92.
Craftsmanship, 197,
Crosses, 149.
Cruciform domed church, 71, 78.
Crusaders, 196.
Ctesiphon, 24, 109.
Cufic lettering, 106.
Cushion capital, 81, 85, 95, 146.
Cyzicus, quarries of, 146.
Czechs, 62.
Damascus, great mosque of, 107.
Daniel, 160.
Daphni, 62.
Dara, 147, 166.
Dashlut, 175.
David and Goliath, 160.
Dawn clouds, 151.
Decoration, 209.
— of surfaces, 148.
— painted, 160.
Dehio and Bezold, 231.
Deir-es-Siiryani, 114, 240.
Diarbekr, see Amida.
Didactic art, 163 ff.
INDEX
263
Diocletian, 73.
— palace of, 145.
Diptychs, 165, 167, 168.
Dobbert, E., 3.
Dome, 8, 22, 27, 57 ff., 86, 202.
— corbelling of, 59.
— on squinches, iii.
— over square plan, 55, 57, 59, 69, iii.
— decoration of, 133 ff.
— of the Rock, The, 106.
Donatello, 185.
Door, 147.
Dragon, Persian, 125.
Duccio, 170.
Durer, 175, 185.
Durrow, Book of, 232.
Dwin, Council of, 37.
Earls Barton, 234.
Easter prayers, 159.
Edessa, 3, 6, 15, 24, 32, 41, 42, 52,
no, 140, 147, 156, 162, 166, 169.
Egypt. 38, 40. 44. 55. 56, 107, 114, 152,
'53-
Khargeh, 137.
Embroidery, Silk, 126.
England, 230 ff.
Ephesus, 40, 132.
Ephraim the Syrian, 169, 176, 177, 180.
Ereruk, 69.
Eusebian canons, 135, 167, 237.
Eusebius, 136, 175, 247.
Facade, 213 ff.
Falke, Otto v., 152.
Felix, Pope, 179.
Ferghana, 228.
Fire temple, 117,
Firuz Abad, 24, 46, 59, 60.
Fischer, Otto, 151.
Fish-bird initials, 127, 241.
Florence, Cathedral, 98.
Form, 206 ff., 212 ff.
Franks casket, 238.
Fundi, Basilica at, 177.
Gagik, King, 159.
Galla Placidia, Mausoleum of (SS.
Nazaro e Celso), 91, 134, 135, 137,
150, 162, 176, 200.
GandMra sculptures, 157, 165, 166,
186.
Gate, 147, 148, 17s, 215.
Gaza, 56.
Georgia, 62, 63, 66, 70, 71, 123.
Germany, 92.
Gesii, the, 73, 100, 191, 219.
Ghasna, 228.
Giorgone, 185.
Giotto, 170.
Goethe, 185.
Goliath, David and, 160.
Good Shepherd, 137, 158, 163, 178.
' Gothic ', 216.
' Gothic' architecture, 12, 21, 55, 72,
86, 93 ff., 154, 187, 197.
' Gothic ' sculpture, 96.
Goths, The, 18, 77, 81, 85, 184.
Greek-Cross plan, see Cruciform domed
church.
Greenstead, 234.
Gregory of Honentz, Church of, 144.
Gregory the Illuminator, 61.
Gregory of Nazianzus, 245.
Gromed vault, 90, 91.
Hall church, 69, 72, 73, 88.
Handicraft, 193.
Hartlepool, 237.
Hatra, 67, 109, 122.
Hauran, 58 ff.
Heavenly City, 171.
Hellenism, 2, 5, 9, 13, 52.
— Byzantine, 16.
Hellenistic culture, 157.
Heracles Maximianus, 175.
Herat, 228.
Hexham, 231, 234, 235, 247.
Hildebrand, 214.
Hindus, 31.
Holy Apostles, Church of, Constanti-
nople, 46, 60.
Holy Sepulchre, Church of, Jerusalem,
83,91, 209, 224.
Horns, 175.
House, Mazdean, 124.
Hripsimeh, 134.
Hunting Scenes, see Chase.
Hvarenah, 118, 160.
— Landscape, 118 ff., 184.
— Symbols, 120 ff., 124, 125, 126, 127,
>33. 147. 159. 203.
264
INDEX
Iconography, didactic, 170.
Ikons, 143.
Illuminated MSS., 126.
Impost capital, 147.
India, 8, 23, 77, 78, 97, 101, 166, 186.
Indian art, 158, 204.
Indogermanic art, 127.
Intercession, 222.
lona, 248.
Iran, 8, 14, 15, 21, 31, 33, 52, 57, 108,
no, 127, 131, 175, 194, 202.
Iranian dwelling-house, 60.
Iranians, 2, II, 31, 188.
Ireland, 230.
Irind, 141.
Islam, 15, 32, 94, 127, 196.
Islamic art, 18, 22, 32, 95, 103, 105 ff.
Ispahan, 147.
Ivory carvings, 165.
Jaxartes, 228.
Jerusalem, 5, 39, 44, 55, 56, 83, 91, 168,
171, 195, 200.
Jesse, Tree of, 160.
Jews, 31, 32, 131, 156, 186, 188.
John the philosopher, 143.
Jonah, story of, 159.
Joshua Roll, 158.
Justinian, 45, 46.
Kabul, 228.
Kaghankatuk, 142.
Kairwan, 44, 125, 241.
Karin, Council of, 143.
Kashmir, 8, 59, 81, 147.
Kells, Book of, 232, 246.
Khakh, 68, 69, 90.
Khilani, Hittite, 215.
Khoja Kalessi, 70.
Khotan, 152.
Kioto, 181.
Klimakos, Ladder of John, 160.
Klinger, Max, 158.
' Knob ' Capital, 146.
Korea, 59, 122, 124.
Kraus, F. X., 5.
Ku-K'ai-Chi, 242.
La T^ne period, 248, 249.
Ladder of John KHmakos, 160.
Lamb, 138.
Lambrequin, 125, 134.
Landscapes, 149, 150, 179.
Landscape, Hvarenah, 118 fF., 148, 149
— River, 120, 136 fF.
Last Judgement, 160, 167, 176.
Last Things, 176 ff.
Lateran Baptistery, 136, 150.
Leonardo, 64, 67, 98, 99, 100, 185.
Le Puy, Notre-Dame, 90.
Libellus Islandorum, 250.
Light, 219 ff.
Lindisfarne Gospels, 232, 235, 246, 248.
Lombards, 18, 81, 85, 88, 184.
Lydda, 175.
Magistri comtnacini, 196.
Manazkert, Council of, 143.
Manichaean art, 126.
Manichaeans, 19, 23, 90.
Margiana, 227, 228.
Marmara, islands of, 146.
Mass, 213 fF.
Mastara, Church of, 61, 134.
Materials, 194.
Maximianus, chair of, 113, 151, 165.
Mazdaism, 17, 19, 27, 115 fF.
Mazdean art, 115, 116.
— costume, 125.
— house, 124.
— ideas, 171, 182.
Mediaeval art, 163.
Meiafarqin, 70, 147.
Meriamlik, 71.
Merovingian MSS., 127.
Merv, 24, 228.
Mesopotamia, 6, 7, 22, 27, 32, 33, 58,
62, 70, 108, 166, 167, 194.
Metzger-Milloue, 166.
Michelangelo, 185.
Michelozzo, 98.
Milan, 41,77, 84, 195,209.
Minerva Medica, at Rome, 65.
Mithraic reliefs, 158.
Mithraism, 115.
Moissac, 185.
Monotheism and art, 104.
Monkwearmouth, 239.
Mokvi, 71.
Montmajeur, 65.
Mosaics, 133 fF., 171, 176 fF., 186.
Mosque, The, 106, 148.
Mount Athos, 161.
fl
INDEX
265
Mshatta, 22, 113, 122, 123, 134, 147,
151,239.
Miiller, Sophus, 154.
Musmiyeh (Mismiyeh), 64.
Mzchet, 242, 245.
Nabataeans, 31, 161.
Naksh-i-Rustam, 117.
Naples, 134.
Nara, 120.
Nativity, Church of, at Bethlehem, 163,
183, 205.
Nave, transverse, 8, 58, 67, 68, 70.
Nea, The, 71, 170.
Nea Moni, Chios, 62.
Nestorians, 23, 25, 132, 169, 228.
Niche-buttress, 61.
Nigg Stone, 240.
Nilus, 84, 136, 144, 150, 163, 187.
Nisibis, 6, 24, 26, 32, 41, 51, 131, 132,
140, 156, 162, 169, 206, 249.
Nomadic life, influence of, 102, 103, 124.
Non-representational art, 18, 32, 42,
102 ff., 130 fF., 202.
Non-representational reUgions, 128.
North, Art of the, 13, 81, 104, 105, iii,
130. 153. 154-
Object, 199.
Octagonal plan, 58.
Odzun, 237.
Olympiodorus, 144, 164.
Orientation, 78, 116, 201.
Orthodox Church, 170, 202.
Oseberg Ship, the, 114, 153, 197.
Oxus, 228.
Padmapani, 123.
Painters, 142.
Painters' Manual, 169, 170.
Paintings, 143, 166.
Palermo, 235.
Palestine, 56. See also Bethlehem,
Jerusalem.
Palmyra, 166.
Pantokrator, Christ, 161, 162.
Papyrus, 157.
Parchment, 157.
Parenzo, 183.
Parthia, 2, 227.
Patleina, 114.
PauHnus of Nola, 137, 177.
Paulus, 136.
Pendentive, 59, 89.
Perigord, 89.
Perigueux, 60, 89.
Persecutions, 53.
Philippi, 71.
Philippopolis, 65.
Pirdop, 71,
Pisa, 176.
Pisano, 170.
Plants, symbolic, 122, 137.
Plato, 182.
Pola casket, 165.
Pompeii, 109, 175.
Portraiture, 156.
Proconnesos, 10, 44.
Psalter 139 in Pans, 158.
putti, 129.
Pyxis, Antiochene, 172.
Qasr Ibn-Wardan, 70.
Quseir 'Amra, 119, 122, 135, 139.
Rabula, Gospels of, 135, 167, 237.
Rams' heads, 147.
Raphael, 185.
Ravenna, 38, 40, 45, 61, 82, 84, 85, 91,
124, 134, 136, 163, 171, 195, 209.
— Baptistery, 163.
Redemption, 159, 160.
Rembrandt, 185, 199.
Renaissance, The, 75, 97 ff., 170,
Representation, 210 ff^.
Representational art, 14, 18, 22, 32, 41,
127, 155 if., 171, 184,
— decoration, 204.
Resafa, Basilica or, 139.
Rhine, Valley of, 88, 92.
Rhone, Valley of, 88.
Riegl, 193.
Rigveda, 174.
Ripon, 234, 235.
River scenes, 120, 136.
Rivoira, 193,251.
Rock caverns, 166.
Rock sculptures, 117, 119.
Romanesque, 42, 54, 55, 73, 81 fi^., 86,
87, 92, 114, 117, 185.
Rome, 2, 33, 37, 74, 130, 163, 171,
183, 186, 187, 194, 230. See also
Baths, Catacombs, Gesii, S. Costanza,
Minerva Medica.
266
INDEX
Roofs, timber, 8, 55, 58, 76.
Rosi, 135.
Rumanian churches, 160.
Rusafa, 64, 70.
Russia, 70, 160.
Ruth well Cross, 231, 241 ff.
Sachau, 228.
S. ApoUinare in Classe, 137, 186.
Nuovo, 177, 183.
S. Aquilino (Milan), 149.
SS. Cosmas and Damian, 178 ff., 218.
S. Costanza, Rome, 56, 133, 134, 135.
S. Demetrius, Salonika, 223.
S. Front, Perigueux, 60, 89, 90.
S. George, 174.
Church at Salonika, 135, 163.
S. Germigny-des-Pr6s, 64, 83, 99.
S. Giovanni in fonte, 134.
S. Giovanni in Laterano, 136, 171.
S. Irene, Constantinople, 71.
S. James, Nisibis, 59.
Salah, 68.
S. Lorenzo, Milan, 65, 82, 84, 202.
in Lucina, 179.
S. Luke of Stiris, Phocis, 62.
S. Maria in Capitolio, Cologne, 92.
S. Maria Maggiore, 133, 136, 137, 168,
171, 179, 183, 205, 206.
S. Mark's, Venice, 60, 89.
St. Peter, 202.
S. Peter's, Rome, 3, 73, 76, 99.
S. Pietro, Spoleto, 121.
S. Prassede, 172.
S. Prisco, Capua, 151.
S. Pudenziana, 171.
S. Satiro, Milan, 64, 99.
SS. Seconda and Rufina, 136.
S. Sergius, Gaza, 137.
SS. Sergius and Bacchus, Constanti-
nople, 46.
S. Sophia, Constantinople, 46, 62, 66,
68, 70, 93, 135, 224.
S. Sophia, Salonika, 135.
S. Sophia, Trebizond, 160.
2^'S. Vitale, Ravenna, 45, 46, 47, 66, 136,
183,200,202,205,206,218,221.
Saint, The Mounted, 173 ff.
Sakastene, 228.
Salin, 153, 154.
Salonika, 135.
Samarkand, 228.
Samarra, 23, 107, no.
Samson, 160.
Sanchi stQpa, 157, 219.
Sarcophagi, 159, 163, 165, 169, 211, 241.
Sarnath, Stupa at, 23, 114, 125.
Sarvistan, 24, 59, 64.
Sassanian dynasty, 17.
— rock sculpture, 117, 119, 174.
— silver dishes, 119.
Satius, 136.
Scandinavia, 18, 78 ff., 93, 114.
Schmarsow, 193.
Scotland, 230.
Scroll, Geometric, 112.
Seleucia, 109.
Semites, 2, 7, 11, 156 ; see also Arabs,
Aramaeans, Jews.
Semitic art, 31, 32, 34, 162.
— realism, 161.
Semper, Gottfried, 197.
Shaft, engaged, 62, 95.
Shapur, 25.
Shepherd, The Good, (q.v.).
Shosoin, 120.
Sidamara sarcophagus, 158.
Siegfried myth, 153.
Significance, 199.
Silk Buddhist picture, 181.
Silks, figured, 125.
Silver dishes, 119, 126.
Singer, Charles, 250.
Sofia, 91.
S6-6rt, Chronicle of, 25, 140.
SorCuq, 120, 133.
South, art of the, 158.
Space, 216 ff.
Spain, 87, 88, 134.
Spalato, 39.
Squinch, 59, in, 113.
Stags, 10.
Stein, Sir Aurel, 228.
Stephanus, 136.
Steyr, House at, 96.
Stucco ornament, no, 114, 124, 160.
Stuffs, 152.
Stupa, 149, 157, 158.
' Styles ', 196.
Sun-god, 149.
Support, the (arch.), 145 ff.
Susa, 117.
Suweida, 174, 175.
INDEX
267
Swastika, 125.
Sybel, L. von, 3, 129, 193.
Symbolic art, 14, 34, 120 fF., 138.
Symbolism, 157, 158, 204.
Syria, 9, 26, 39, 40, 55, 56, 57, 58, 67.
Syrian Bible, 177.
Tak-i Bostan, 119, 123, 125, 173.
Tak-i Kisra, 67, no.
Tak Ivan, 107.
Tent, The, 124.
Ter-Mkrttchian, 143.
Teutonic ornament. Early, 102, 103.
Textiles, 152.
Thalin, 70,91, 93, 134, 141,
Thalish, 72, 73, 90.
Theodore, 249.
Theodoric, Tomb of, 61, 82.
Theodosian capitals, 146.
Thomas Artsruni, 155.
Timbered roof, 8, 55, 58, 76.
Tomb, the, 200.
Topdschian, 142.
Torcello, 176.
Transverse nave, 8.
Trdat II, King, 6i.
Trdat, architect, 62, 72, 94.
Trebizond, 160.
Tree, formal, 122, 123, 133, 152,
Trefoil (Trikonchos), 69, 70.
Trichora, cella, 70.
Triforium, 96.
Tulun, Mosque of, 107.
Tunhuang, 152.
Tur 'Abdin, 68, 139.
Turkestan, Chinese, 19, 23, 119, 123,
124, 152.
Turks, 40, 103, 112.
Ukraine, Wooden churches of, 18, 46,
60,81, 89, 147.
Urnas, 146, 153, 197.
Ursus, 82, 135.
Vagharshapat, 37, 142.
Valdres, 80.
Van, Lake, see Achthamar.
Vasari, 215.
Vaulted construction, 8, 9, 10, 22, 55,
57. 76.
Vignola, 73, 98, 100, 191, 202, 219.
Vine ornament, 112, 113, 122.
Vine scrolls, 150.
Vivarium, 249.
Vladimir, Church of, 160.
Vogiie, de, 9, 94, 231.
Wahan, 143.
Wall-decoration, 22, 105, 106-8, 124.
— Tiles, 114.
Wheel, 158.
Wickhoff, 4, 193.
Wilpert, 178, 179.
Wood, see Timber.
Wooden churches, 8, 18, 60, 78 fF., 93,
"6, 153.
Work, 194, 196.
Wu-Tao-Tse, 181.
Wiirzburg, 66, 92.
Yima, 121, 158, 178.
YuriyefF Polskiy, i6o.
Zagba, 167, 177.
Zarathusht-Nameh, 174.
Zarathustra, n6, 138, 200.
Zend Avesta, 17, 19, ii8, 174, 176, 177,
180.
Zenobius, 224.
Zenrinji, 181.
Zimmerman, 179.
Zoroaster. iSee Zarathustra.
Zoroastrianism, 172, 173, 182 ; see
also Mazdaism.
Zwarthnotz, 65, 113, I22, 141, 144, 147,
202.
PRINTED IN ENGIAND
AT OXFORD UNIVERSITY PRESS
PLEASE DO NOT REMOVE
CARDS OR SLIPS FROM THIS POCKET
UNIVERSITY OF TORONTO LIBRARY
' NA
4800
S72
J
1
1
!
Strzygowski, Josef
Origin of Christian chvirch
art
Live Music
Archive
Librivox
Free Audio
Metropolitan Museum
Cleveland
Museum of Art
Internet
Arcade
Console Living Room
Open Library
American
Libraries
TV News
Understanding
9/11